Biomea Fusion, Inc. (BMEA) Earnings Call Transcript & Summary

Good day, and thank you for standing by. Welcome to the Biomea Fusion BMF650 GLP 131 Phase I top line results conference call. [Operator Instructions]. Please be advised that today's conference is being recorded. [Operator Instructions]. I would now like to hand the conference over to your speaker today, Ramses Erdtmann, Co-Founder, Chief Operating Officer and President of Biomea Fusion.

Thank you, operator, and welcome, everyone, to this conference call. Today, we will discuss top line results from the Phase Ib, a multiple ascending dose study evaluating BMF 650 and our investigational oral small molecule GLP-1 receptor agonist, which we developed alongside our menu inhibitor, Ecovomini. Before we begin, I'd like to remind everyone that today's discussion contains forward-looking statements. I direct your attention to Slide 2 of our presentation and our SEC filings for a full review of the risks and uncertainties that could cause actual results to differ. The development of oral GLP-1 therapies has advanced substantially. Across the field, multiple programs have demonstrated that all incretin therapies can deliver clinically meaningful weight loss and glycemic improvement. At the same time, the category continues to confront a practical challenge. As exposure rises to levels intended to produce meaningful efficacy, gastrointestinal tolerability frequently becomes the limiting factor. The challenge is straightforward to describe, but has been difficult to solve. For example, in a real-world 2025 study of more than 125,000 overweight or obese on. who initiated a GLP-1-based therapy, 65% of them discontinued within the first year. Today, no oral program has conclusively demonstrated that potent clinical activity can be fully separated from a meaningful cumulative gastrointestinal burden. Across programs, the most consistent development response has been to primarily adjust and manage the exposure, which includes lower starting doses, slower escalation, additional intermediate steps and a longer time before patients reach maintenance exposure. VMS 650, however, was designed differently. It was designed with the intent to separate weight loss from GI tolerability and thereby allow patients to stay on treatment. With that, let me turn the call over to Thorsten, our Head of Research and leader of the BMS 650 development program. He has overseen this program since inception and will review the design rationale and our early nonclinical results, which provided us a first proof for the emerging 650 profile. Thorsten?

Thank you, Ramses. I will begin with Orfoglepron because it provided an important scientific reference point for our work. Orfoglipron established that a potent orally available non-peptide small molecule can activate the GLP-1 receptor and produce meaningful clinical activity. We admire the rigor and success of that program and the insights it delivered. That brings me to the central objective behind BMF 650, which is captured in the title on Slide 8. The potential to separate GI side effects from weight loss in the oral GLP-1 receptor agonist space. A successful oral GLP-1 receptor agonist should capture the convenience of oral dosing, while addressing some of the limitations of the class. With BMS 650, we specifically designed the molecule for less acute GLP-1 receptor agonism in the GI tract with the goal of improving GI tolerability. If successful, that could enable fewer titration steps ideally just one, and ultimately support greater persistence with chronic therapy. Slide 9 takes you into the molecule design behind that hypothesis. BMF 650 is a selective and biased GLP-1 receptor agonist with no off-target concerns identified in our counter-screening assets. Importantly, BMS 650 has approximately 26-fold lower agonist potency than orforglifron. This reduced potency is an intentional element of the design and part of hypothesis for reducing acute GLP-1 receptor activation in the GI tract. Slide 10 shows another aspect of the design, bioavailability. In nonclinical PK studies in both monkeys and rats BMF 650 demonstrated high and prolonged exposure following oral dosing. Importantly, bioavailability was approximately 2 to 3 higher than orforglifron in these experiments. So conceptually, the design is quite straightforward. Lower receptor potency coupled with high and sustained systemic exposure. The long clinical PK gave us confidence to take that hypothesis into humans, and you would see the human PK data in the clinical section that follows. But before entering the clinic, we wanted to know whether this profile could translate into meaningful weight loss. We, therefore, evaluated BMF 650 in obese sunomorbous monkeys using once-daily oral dosing. As shown on Slide 11, we observed significant dose-dependent weight loss of approximately 12% and 15%. The response was consistent across animals within the different treatment groups and showed no apparent attenuation towards the end of the study. Importantly, any events were relatively limited despite dosing without titration. On Slide 12, I bring together these nonclinical observations to highlight 3 key elements of our design. First, BMF 650 uses a relatively higher oral dose and was designed for broader intestine absorption with the intent of reducing localized concentration peaks within the GI tract and supporting absorption across a broader intestinal surface and time window. Second, BMF-650 demonstrated high order bioavailability with low variability in our nonclinical studies. A hypothesis was that this should translate into consistent and predictable exposure in humans. And third, BMF 650 was designed to produce a gradual increase in systemic exposure with a relatively low peak to trough ratio providing a more even exposure profile over the dosing in Durban. Taken together, this was the design hypothesis going into the clinic. Lower acute receptor activation combined with sustained systemic exposure could potentially achieve meaningful weight loss activity while improving GI tolerability. We sometimes describe the intended distinction as a blow torch versus a radiator, concentrated intensity versus a more gradual and sustained delivery of the biological segment. Against this background, let's now turn to the Phase I clinical data and see how BMS 650 behaves in humans. For that, I will turn the call to Dr. Juan Pablo Frias, who leads our Scientific Advisory Board supporting the BMS 650 program. Juan Pablo is a recognized clinical investigator in diabetes, obesity and metabolic disease with extensive experience evaluating incretin-based therapies. Juan Pablo, over to you.

Thank you, Thorsten. Today, I'll review the top line results of the Phase I study of BMF 650 and I will focus on 4 areas: the study design, weight reduction, safety and tolerability and pharmacokinetics. I will then briefly summarize what we believe these findings support as the program moves forward. Let's begin with study design on Slide 14. This was a randomized, double-blind, placebo-controlled Phase I study in overweight or obese participants. The first part evaluated single ascending doses of BMF 650 primarily to assess safety, tolerability, pharmacokinetics, pharmacodynamics and the effect of food. The second part was a 28-day multiple ascending dose study. This part was primarily designed to evaluate co-primary tolerability measures, including adverse events, vital signs, EKG measures, like QTcF interval and safety labs. The secondary endpoints included percent change in body weight at study end along with pharmacokinetic measures. We initially included 4 active dose cohorts and corresponding placebo participants. The dose regimens increased across the cohorts. The first 4 cohorts evaluated different weekly dose escalation schedules. Now let's turn to the baseline characteristics on Slide 15. Overall, demographics and baseline characteristics were reasonably balanced between BMF 650 and pooled placebo groups. Participants had a mean BMI of approximately 34 in the active group and approximately 36 in the pooled placebo group. Mean baseline body weight was approximately 100 kilograms in the active group and 110 kilograms in the poor placebo group. Each cohort included 8 active and 2 placebo participants for a total of 32 active participants treated with BMF 650 and and 8 pool placebo participants in the first 4 cohorts. Turning to the clinical activity findings on Slide 16. This analysis examines the incremental weight change during each weekly dosing interval of cohorts 1 through 4. Here, we observed reductions across all 4 BMS 650 treatment cohorts through day 29 compared with essentially no change in pool placebo, mean reductions range from 1.5% to 3.2%, with cohort 3 achieving the greatest mean reduction of 3.2% in this key secondary end point. While these are short duration Phase I data from small cohorts, the consistent direction of weight reduction across the active cohorts provides encouraging early evidence of clinical activity for BMF 650. On the next slide, you can see that across these first 4 cohorts, 14 of the 16 weekly dosing observation showed additional weekly weight reduction and the magnitude of the weekly weight reduction generally increased at higher administered doses. Across these cohorts, the mean weekly weight reduction at doses of 200 milligrams once daily or higher, was approximately 1 kilogram per week. In Cohort 4, we observed an increase in weight during the first week, which was unexpected. Closer respection of the pharmacokinetic data raise concern of some participants may not have taken the study drug consistently. We, therefore, introduced Cohort 5 under confinement to ensure dosing compliance. We also evaluated a simpler titration schedule in this additional cohort starting at 200 milligrams once daily and increasing after 1 week to 400 milligrams once daily for the remaining 3 weeks. At that point, we had observed only limited safety findings and importantly, no clear increase in gastrointestinal adverse events in the earlier cohort. Turning now to the results of this additional cohort 5 on Slide 20, we can see that the confinement allowed close assessment of safety and tolerability, but it also appears to have influenced eating behavior. All participants were provided 2,100 calories per day. The 2 placebo participants consumed approximately 33% of the presented food and may have been sensitive to the food choices. They experienced weight reduction similar to the active participants, which complicates the interpretation of weight loss under confinement. Cohort 5 participants were CDMS 650 lost an average of approximately 1.55 kilograms per week for a total weight reduction of 6.2 kilograms or 6.1% after 4 weeks with no observed attenuation over the treatment period. The case cohort 5 was small and the 2 placebo participants consumed less food than active participants, the results of this cohort should be interpreted cautiously. We, therefore, view Cohort 5 as an exploratory analysis and consider it together with the dose response pattern across the complete data set the pharmacokinetic findings and the observed tolerability profile. Taken together, these findings support potential meaningful pharmacologic activity. Turning to safety and tolerability on Slide 21. Overall, BMF 650 was generally well tolerated across the 5 active dose cohorts. All treatment-emergent adverse events were mild or moderate, no severe or serious treatment-emergent adverse events were reported. There were no dose modifications and no study discontinuations related to treatment-emergent adverse events. We also did not observe a clear association between treatment-emergent adverse events and dose escalation, dosing cohort or time at a given dose. On the next slide, we show selected gastrointestinal adverse events. Across the 40 participants treated with BMF 650 in the MAD cohort, nausea was reported in 7 participants, constipation in 4, vomiting in 3 and diarrhea and 2. In the highest dose cohort in the selected gastrointestinal events were all observed to be grade 1 in severity. Only 1 of 8 participants in Cohort 5 reported nausea and 1 reported a single vomiting event. There were no reports of diarrhea in this cohort to participants reported constipation. Slide 23 provides post-hoc descriptive context relative to selected Phase I disclosures of other oral GLP-1 receptor agonists that have progressed to late-stage development. At the highest BMF 650 dose regimen, the observed incidences were 12.5% for nausea, 12.5% for vomiting, 0 for diarrhea and 25% for constipation, no head-to-head studies were conducted, and these cross-trial comparisons should be viewed cautiously. The study is differed in size, duration, dose escalation schedules and adverse event collection and reporting. Nevertheless, we believe the observed gastrointestinal profile of BMF 650 supports further evaluation at the 400-milligram dose. Slide 24 shows selected changes in liver enzymes, fasting glucose and heart rate. Mean ALT and AST values decreased from baseline in both the active and placebo groups and no adverse events involving ALT or AST elevations were reported. Fasting glucose decreased across all BMF 650 dose cohorts with a mean reduction of 9.2 milligrams per deciliter in the pooled active group compared with 3.5 milligrams per deciliter in pool placebo. Mean heart rate increased by 6.4 beats per minute in pooled active group compared with 0.8 beats per minute in pooled placebo, no adverse events involving heart rate were reporting. Turning to pharmacokinetics on Slide 25. And which were key secondary measures, BMF 650 achieved the target exposure predicted from the nonclinical studies in cohort 5 on day 28 and the mean exposure over 24 hours with approximately 79,000 hour nanograms per milliliter. The observed peak per trough ratio was approximately 1.7%, indicating relatively limited fluctuation in exposure over the 24-hour dosing interval. The observed human exposure was also consistent with the efficacious exposure observed in the relevant nonclinical weight-loss study. The pharmacokinetic analysis remains preliminary but these results support once daily administration and confirm the clinically relevant exposure was achieved. Turning to Slide 26. In summary, BMF 650 was designed to separate weight reduction from GI safety risks. Across the 5 active cohort, we observed that with increasing dose and exposure, participants achieved greater weight reduction while maintaining a very well-tolerated profile with comparatively low GI effects. Bringing these findings together, the Phase I results support the original BMF 650 design hypothesis in 3 important respects. First, participants reached the target dose of 400 milligrams once daily through a single titration step after week 1, supporting a potentially simple and practical dosing paradigm. Second, Cohort 5 demonstrates sustained and meaningful weekly weight reduction of 400 milligrams across 3 consecutive weeks, reaching mean weight reduction at day 29 with no evidence of attenuation over the observed treatment period. And third, the GI tolerability profile was very encouraging across the active multiple ascending dose cohorts, there were no serious treatment emergent adverse events, no treatment-emergent adverse events leading to dose modification and no treatment-emergent adverse events leading to study discontinuation. Biomea plans now to continue with longer-term toxicology studies to support Phase II clinical evaluation and advance the Phase II planning. -- including the assessment of whether participants can initiate treatment directly at the highest dose without titration at all. With that, I will turn the call over to Mick, the company's interim CEO.

Thank you, Juan. I am very pleased with the results BMF-650 demonstrated in this first clinical trial. In the highest dose cohort presented today, we observed meaningful weight reduction, together with an emerging safety and tolerability profile that we believe supports the design objectives we established for BMF 650. Our goal was to create an oral GLP-1 receptor agonist while modifying its properties to improve tolerability and support sustained use over time. When we look at BMF-650 in the context of the Phase I results reported for other advanced oral GLP-1 receptor agonists. We believe the combination of clinical activity and tolerability data we observed is encouraging. Importantly, we are not looking at the clinical activity or tolerability data in isolation. We believe the opportunity is to deliver both profiles and ultimately to provide a treatment experience that can support long-term use. Based on the results presented today, our intention is to continue advancing BMF-650 towards Phase 2 development. The next steps include completing the additional nonclinical work required to support longer duration dosing, manufacturing clinical supply and preparing the program for the next stage of clinical development. At the same time, I want to be clear about how we think about capital allocation. We intend to invest in BMF-650 through the stages where we believe we can further establish its clinical profile and maximize the strategic value of the program. We do not currently envisage Biomea independently funding a large global clinical obesity program. Our objective is to advance the BMF-650 and position the program for a strategic partnership, which could include a licensing arrangement for late-stage development and commercialization. We believe this strategy allows us to continue building the value of BMF-650 while remaining disciplined in how we deploy Biomea's capital. At the same time, icovamenib remains at the core of Biomea and represents a major potential value driver for the company. Our focus now is on validating the highly encouraging clinical signals we have observed to date and advancing the program through a series of important milestones over the next 18 months. Over the next 2 quarters, we plan to initiate 2 Phase II studies in type 1 diabetes. Here, we are expecting to get a first set of 6 months response data from 36 type 1 patients within the next 18 months. Then in the first half of next year, we expect 2 additional important Phase II readouts. The 26-week primary endpoint for icovamenib in insulin deficient patients, as well as the primary endpoint from our study in patients with inadequate glycemic control on GLP-1 therapy. And towards the end of next year, we also expect results from our study evaluating icovamenib in combination with semaglutide, where we are investigating the potential for additional weight loss together with preservation of muscle bone and physical health. Taken together, these milestones are designed to help us organize end of Phase II meetings with the FDA for type 2 as well as type 1 diabetes and move icovamenib to late-stage development. We thank you for your continued support and appreciate your interest in the work we are doing at Biomea. Operator, we are now ready to take questions.

[Operator Instructions]. And our first question comes from Jon Wolleben with Citizens.

Congrats Biomea. Wondering if you could talk a little bit about the timing to complete a tox study. And then it sounded like you guys would be open to running a Phase II program. wondering I could think about some initial thoughts given the freshness of this data? And then maybe a question for Dr. Frias. Wondering how you think about or differentiation and how many oral drugs the field may need and what actually constitutes a clear differentiation between these molecules.

Jon, it's Mick. In terms of the Phase II, I think given where we came out on the Phase I, we want to do another dose in the Phase I, possibly 400 directly and figure out what was going on in the short term with that. that would lead us then to potentially a Phase II design where we would have no uptitration. We just started the dose that we were going to use and see how well that works over time. I'll pass the other question over to Juan.

Yes. So I think -- I mean, the answer is quite similar to what you said. I think the differentiators would be, one, the tolerability of this pans out as we look further in the clinical development program. but then also either having a very short escalation, basically one dose step or the tolerability is there to be able to initiate and just have one dose. So I think that would be a fairly important differentiating feature. So I think there certainly is room for a drug that has those characteristics.

What's the minimal amount of weight loss you want to see for a drug to be worthwhile to use?

I think probably somewhere in the range of 10%. I think would be -- it could even be a little lower than that. But I think 10% is still on average, very good weight loss for a lot of patients. I would also say that the patients who would benefit from less weight gain, particularly if we're trying to prevent obesity-related complications and other cardiovascular risk factors. But I think somewhere in the range of 10% as a minimal.

Maybe, Jon, I just want to add one more comment because I found it so profound when I looked at the data that in the highest dose when we went into the and where we put the patients up in confines and we added the highest dose with one-step titration. The safety profile was upheld and if you look at the safety profile and the tolerability that the drug has shown, it's so comparatively superior in my mind, and this is a personal view. But I was amazed that if you just do 1 step titration going from 200 to 400 you're not incurring more safety issues. One, it stays fairly benign in comparison to all the other players in the field. So if you were to ask me what is the key differentiator between us potentially as we validate our signals in further studies and others, it could be that the safety is the key point of differentiation. One, we can go faster there. And two, we're just unusually safe based on the design. And we spent quite a few slides on explaining the design and Thorsten section goes into the reason for why we designed the drug we did -- but I would argue that is for me the key differentiator, trying to give somebody weight loss without all the GI tolerability issues that have to go through. Anyway, next answer next question?

And our next question comes from Joe Pantginis with H.C.Wainwright.

Very nice start to the data. So I have a couple of broader questions looking forward here. First, can you give sort of the profile you would be looking at with regard to a BB partner -- and that also links to the profile of the drug you might want to see in the future. Obviously, as Ramses just said, a better safety profile. But would you look at maintenance therapy, frontline therapy, what kind of options would you potentially see for this drug? And -- so that's number one. And then number two, what would you consider to be -- and you touched upon this a bit during your presentation, the open questions with regard to finalizing the dose and the quicker titration.

Yes. Thanks, Joe. So if you look at the field, and that was sort of in my section, persistence is the problem meaning we have a great backbone here as in the GLP-1 strategy on how to address obesity, yes. So we understand that part -- but we haven't fully figured out as a field on how to really do it in a way that we can get patients to stay on drug. And that's why we quoted that recent survey from 2025 in our slide deck. So if 125,000 users, 65% are not staying on drug after -- within the first year, something is wrong. And the way we designed the drug, and that was evident already 5 years ago. the way we designed this compound was to achieve persistence. And that's where I think a partner would find potential entry into the space. If you look at everybody was going for maximum weight loss, and that's great. And now who's staying on drug and who can achieve persistence with patients potentially wins in the long run. Our goal was not to achieve maximum weight loss. Our goal was to achieve a good weight loss over time. You can step into this compound from another drug, you can -- you don't have to uptitrate, you can easily just switch if you're unhappy with what you currently are, those kind of aspects, I think, could potentially add value. And so [indiscernible] partner would be somebody who has a fair program -- it could be somebody who has a program but hasn't gotten this aspect in his own drug or it could be somebody who hasn't entered the space and once the con in the space was something new. We're looking at all 3 of these categories. We're pursuing each one of them. And as a small biotech company that is surviving on good data and also nondilutive funding from pharma. That's where we are seeing opportunities, particularly with 650. Now you asked another question on finding a dose, and Mick alluded to that, we think another cohort 6 in a nonconfined setting, so in an open setting where people have these choices for eating as much as they want to would hopefully -- not hopefully, we think actually that would be a great setting for us to show the key difference between placebo versus us -- we already know the safety because in confinement for sure, the physician sits right next to the patient and Picon every AE profile you may imagine. So safety-wise, we feel really certain that this is what it is. [indiscernible] wise, yes, the placebo, we're a little bit behaving quite unnormal, and that can happen if you have a small end, and it did happen to us here. So what we're thinking of doing is initiating another cohort 6 potentially with a partner. Where we would go in, add another, whatever, 10, 15, 20 patients with a larger placebo setting, not in a confinement situation and run that cohort. That would more or less clarify the signals we have seen and allow somebody else very likely to take this drug into Phase II.

I appreciate that, Ramses. And if I could just set a quick 1 in there. Have any discussions started yet? Or are these data the impetus to go ahead?

So I can't really comment too much on it, but you can imagine that -- we don't want to wait for pharma to approach us. We're doing everything we can with what we have to engage conversations like every company should do in our situation. So you're going to assume we've done the same.

Our next question comes from Cha Cha Yang with Jefferies.

I was just wondering if you can speak a little bit to the dose response that we see or specifically about Cohort 3 performing better than cohort of? Do you think it's mostly a small end issue? Or what other things are we thinking about here?

Cha Cha, this is Mick. I think the -- the thing that's important to see, and I mean, this is the reason we've broken it out is the dose response is pretty good if you get rid of that first point in cohort 4. So we think something went wrong there, which is basically changing the starting point for the rest of the line to go down and I think, again, it is small numbers. There's only 8 in each active cohort. And so that probably has something to do with it as well. But I really think it's the fact that it went up on the first dose -- sorry, on the first week, and that sort of changed the starting point for the line to come down. Yes, anything else to add Thorsten.

When you look at the cohort from compliance. Yes. So it is true the Cohort 4 generated the most question marks towards compliance because the PK, as we've shown, is very reliant when you can measure it for participants. So the -- essentially, the 0 value on certain days is a good indicator of drug compliance on the day before and so some of the analysis we did cohort 4 points towards the main problem along those lines. But patients didn't take the drug because they didn't have sufficient exposure, right and you would have expected more exposure. And that kind of led us to the conclusion we get into Cohort 5 because some patients haven't taken the drug. It would be force to conclude that Cohort 4 is reality, right? And so that's why we said, okay, that Cohort 5, if you put a mix to each other is obviously under confinement. When everybody takes the drug is a beautiful response line. But we think a cohort 6 would in a nonconfined setting will probably help clarify that.

Our next question comes from Yigal Nochomovitz with Citigroup.

So I had a few here. The first one is when you look at the NHP data, which showed the 15% reduction at a month, can you just explain a little bit more what the -- how the exposure on the AUC and the Cmax compared to control -- in those animals versus Cohort 5 in the confinement experiment. Were those similar? Or was the exposure higher in the monkeys? And then can you just explain a little bit more on the consignment. These patients were essentially in consignment 24/7 for the entire month? Or was it just at certain points during the day when they were eating or taking the drug?

I'll start with the [indiscernible]. So with respect to the exposure, the monkey starting the 400-milligram titration target of these Cohort 4 or Cohort 5, they align very well. Actually, the exposure that you see as the area thousand on day 28 in cohort 5. This is essentially what the high dose cohort of the monkeys had a daily exposure. And then it's approximately 2 to 3/4 less in the monkey for the low dose cohort, which aligns approximately between 200 and 300 milligrams titration target of Cohort 2 and Cohort 3. So everything kind of came out nicely. And personally, I'm actually a fan of cohort 3. The weight loss, placebo-adjusted reemit almost off of [indiscernible] actually. So that's that. And then addressing your question about the confinement. So cohort 5 was a complete confinement study for 30 days actually in the unit. So participants didn't leave the unit, everybody had their room and it was kind of staying there sleeping. They are taking them near and supervision, getting the protocol assessment and dosing on the supervision for the over time.

Okay. So do you have a hypothesis as to why in the animals, you saw mid-teens percent versus what looks like about -- I guess, 6% in the cohort 5? Is that -- how do you reconcile those?

Yes. So thank you for the question. So for us, as a medicinal chemist, I always faced with translating animal data to human data. And every animal almost has a conversion factor for PK parameters. I apply these for almost everything else. So for monkey general, that's 3. So if you divide what you see in the monkey by 3, you actually almost really get to the human human numbers. So I would think a monkey with high metabolic grade and the smaller species, altogether more hardware, everything this comes together to just adjusting it downward by approximately this factor. So I'm actually not surprise. In fact, that's what I would have expected.

Our next question comes from Michael King with Rodman & Renshaw.