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

Ladies and gentlemen, thank you for standing by. We do apologize for the delay due to technical difficulties, and we welcome you to Biomea Fusion COVALENT-111 Topline Results. [Operator Instructions] Please be advised that today's conference is being recorded. I would like now to turn the conference over to Ramses Erdtmann, President and COO. Please go ahead, sir.

Thank you very much, operator. I apologize for the delay. I'm really glad we got it started now. Welcome, everyone, to this conference call. Today, we'll be discussing the topline results from the COVALENT-111 study, a randomized, double-blinded, placebo-controlled Phase II trial to assess the efficacy, safety and tolerability of icovamenib in patients with type 2 diabetes. I'm joined today by Thomas Butler, Chair and Chief Executive; and Dr. Juan Pablo Frias, Chief Medical Officer. We are also pleased to have with us today, Dr. Alex Abitbol, he will join us. He is the principal investigator from our study and a member of our Scientific Advisory Board. The call will begin with opening remarks from Tom, who will turn it over to Juan, who will walk us through the results, then we will ask Dr. Abitbol to give his perspective and insights on the data before turning it back to Tom to close out the presentation. After the prepared remarks, the presenters will be joined by Dr. Steve Morris, Chief Development Officer; and Franco Valle, Chief Financial Officer; and also Thorsten Kirschberg, who is Head of Chemistry for the Q&A session. The slides that will be reviewed today can also be found in the Events and Presentations section on our website. Before we start, let me remind you that this nonconfidential presentation contains forward-looking statements about the business prospects of Biomea Fusion. These statements are subject to a number of risks and uncertainties that could cause our actual results to differ materially from those expressed or implied in this presentation depending on the progress of Biomea's preclinical and clinical development activities, actions of regulatory authorities, availability of capital, future actions in the pharmaceutical market and developments by competitors and those factors detailed in Biomea's filings with the SEC, such as 10-Q, 10-K and 8-K reports. All forward-looking statements made during this presentation are based on the beliefs of Biomea as of this date only and future events or simply the passage of time may cause these beliefs to change. Please be aware that you should not place undue reliance on the forward-looking statements made today. Okay. I will now turn the call over to Tom Butler. Tom?

Thank you, Ramses. Good morning, and welcome to the call, everyone. On behalf of our Biomea team, I am incredibly gratified that our COVALENT-111 study in patients with type 2 diabetes is a win. The outcome today exceeds our expectations, and it delivers the target product profile we had hoped to achieve when we started this exploratory Phase II study. This topline readout provides a clinical and statistical outcome that provides proof of concepts validating covalent menin inhibition in diabetes as a potential new treatment modality. We had seen in our animal experiments that we are increasing beta cell mass and function. We also saw a similar impact in human donor islet cell experiments. And we can now conclude that we see the same impact in the clinic in patients suffering from type 2 diabetes. The data have shown that icovamenib delivered clinically significant reduction in hemoglobin A1c in our target patient population, which represents over half of the U.S. type 2 diabetes population. The results presented today are exceeding our expectations. With this study readout, we have validated that an oral covalent menin inhibitor has a profound impact on diabetes patients. And even more importantly, we have demonstrated that our covalent menin inhibitor, icovamenib, was generally very well tolerated. Our results are in line with the postulated mechanism of action. We now understand from this data set who are likely to be the best responders to icovamenib. The data we presented today is a clear win. We now have a defined path to developing this molecule in diabetes. These data also provide us a robust and strong package we can now use to discuss next steps with the FDA. We believe this is a truly landmark event for Biomea and diabetes patients. Icovamenib has led to a statistically significant reduction in A1c in all study participants who are suboptimally controlled on one or more antidiabetic agents at baseline and who completed their study drug dosing per protocol, but with significantly greater benefit in those that were categorized as having more severe insulin deficiency. This is exactly what all our preclinical data was suggesting. We now have further evidence and could not have asked for a better trial outcome. As a company, we are incredibly proud of the potential to address an aspect of diabetes that no other current therapy can address. All agents used in diabetes care today are primarily addressing glucotoxicity and the ensuing side effects. They are not addressing why patients with diabetes are experiencing such glucotoxicity. The United Kingdom Prospective Diabetes Study revealed in 1998 that at the time of diagnosis, 40% to 50% of beta cell function is already lost with a further decrease in beta cell function as the disease progresses. This is what we are targeting with icovamenib, the depleted pool of beta cells. The majority of complications, including death through heart disease, kidney disease, nerve damage and many others lead to approximately 12 to 14 years of life lost for all those with type 2 diabetes. We are proposing with our study results a new and novel mechanism of action to help curb this trend without placing additional burden on patients. The data released today makes us hopeful that icovamenib can substantial benefit to patients and in the future will become part of the treatment regimen. My sincere thanks go out to the patients who have participated in our studies to whom we owe these results and as well to my colleagues and the investigators who contributed their skill and support to bring us to these results. Before I ask our CMO, Juan, to walk you through the COVALENT-111 data, let me reiterate, today marks a historic moment for Biomea. The results we presented here are an unequivocal success. The data from COVALENT-111 are clinically highly meaningful and statistically significant. We believe this dataset, combined with the favorable safety profile we have now seen in this population, will build the foundation for a clear clinical path development. We intend to meet with the FDA after we have received the final data readout, week 52. In addition and possibly even more significantly, we believe today's data show that covalent menin inhibition can have a meaningful impact on diabetes patients, an effect that has been achieved without chronically dosing patients and addressing the root cause of the disease, insulin deficiency. This lays the foundation of our diabetes pipeline, allowing us to consider an even more broad development icovamenib, not only in those patients that have relapsed on a frontline agent, typically metformin, but also those that are late-stage diabetic or even insulin-dependent. In short, we are very excited with today's results. Let me turn the call over now to Juan, who will walk you through the data.

Okay. Thank you, Tom, and good morning to everyone on the call. It's my pleasure to present the topline results from the expansion phase of COVALENT-111. We are very pleased with the outcome for several reasons. Importantly, as you'll see in the key safety tables I'll present, icovamenib safety profile was extremely clean in the 168 patients who received the drug. Additionally, as we have postulated based on icovamenib's mechanism of action to increase beta cell mass and function, we saw greater and quite robust efficacy in patients in the insulin-deficient subtypes of type 2 diabetes compared with more insulin-resistant participants. And lastly, we were able to validate the icovamenib dose we will take forward into our next clinical trials. So all in all, based on the data we are sharing today, we feel that icovamenib has the potential to be an important future therapy for people with type 2 diabetes, particularly those in which insulin deficiency is a principal pathophysiologic defect leading to hyperglycemia. A very important concept and one that has gained a lot of traction in recent years is the fact that type 2 diabetes is an extremely heterogeneous disease with respect to many factors, including the predominant pathophysiologic defects that lead to hyperglycemia, high blood glucose, which defines diabetes. This quote from the American Diabetes Association Standards of Care states us very nicely, and I read, while diabetes is diagnosed on the basis of a single metabolite, glucose, hyperglycemia can arise due to multiple complex etiologic processes that can vary between individuals. Two key etiologic processes are insulin deficiency, so a reduction in insulin secreting beta cell mass and function and peripheral insulin resistance, the inability of tissues such as muscle and fat to appropriately respond to insulin. And one of the most well-recognized and validated methods by which to classify persons with type 2 diabetes based on these defects is shown here. This cluster analysis methodology that uses key patient features, hemoglobin A1c, measures of insulin deficiency and insulin resistance, HOMA-B and HOMA-IR, body mass index and age places patients into categories that have both prognostic implications with respect to potential future complications and also therapeutic implications with respect to which antidiabetes therapies may offer the best glycemic control, individualization or personalization of therapy. These categories, which I'll come back to later as we discuss the COVALENT-111 efficacy data include severe insulin-deficient diabetes, or SIDD; severe insulin-resistant diabetes, or SIRD; mild obesity-related diabetes, MOD; and mild age-related diabetes, or MARD. The SIDD and MARD categories are characterized by a greater degree of insulin deficiency and, importantly, in this and many other studies have been shown to represent a significant proportion of people with type 2 diabetes, depending on the population assessed anywhere from 40% to 70%. With this as background, let's now move to the COVALENT-111 study design. The expansion phase of COVALENT-111 enrolled adults up to 65 years of age with type 2 diabetes up to 7 years since diagnosis. At study entry, patients were poorly controlled with a hemoglobin A1c between 7% and 10.5% and needed to have a BMI or body mass index between 25 and 40 kilograms per meter squared. So all participants were either overweight or obese, and they were treated with up to 3 antidiabetic agents, excluding insulin and insulin secretagogues. As you can see in the schematic of the design, this was a randomized, double-blind, placebo-controlled 3-arm study. Within each arm, A, B or C, participants were randomized in a 3:1 fashion to icovamenib or placebo. In Arm A, icovamenib 100 milligrams daily was administered for 8 weeks, followed by 4 weeks of placebo. In Arm B, icovamenib 100 milligrams daily was administered for 12 weeks. And in Arm C, icovamenib 100 milligrams daily was administered for 8 weeks, followed by 4 weeks at a dose of 100 milligrams twice daily. So as in Arm B, 12 weeks of icovamenib therapy in Arm C. Importantly, baseline antidiabetic medications, if they were taken, were continued at a stable dose. So 8 or 12 weeks of icovamenib treatment for the active arms with continued follow-up until week 52. The primary endpoint is a change from baseline in hemoglobin A1c at 26 weeks as well as safety and tolerability. These are the endpoints we will be discussing in today's call. The study is continuing, and we will have 52-week data next year. Before we move into the results, I want to provide some context from a post-hoc analysis we conducted in the multiple ascending dose or escalation phase of COVALENT-111 in which patients with type 2 diabetes were treated with 4 weeks of daily icovamenib at various doses and then followed for a subsequent 22 weeks. In this analysis, we looked at the change from baseline in hemoglobin A1c in the insulin-deficient subtypes, MARD and SIDD and in the insulin-resistant subtypes, MARD and SIRD. On the right, you see the results, a 1.23% reduction in hemoglobin A1c in the insulin-deficient patients and a 0.48% reduction in the insulin-resistant subtypes. This greater efficacy in insulin-deficient participants, which make up at least 50% of the type 2 diabetes population as well as the mechanism of action of icovamenib, which is to restore beta cell mass and function, led us to hypothesize that icovamenib may be particularly efficacious in these insulin-deficient type 2 diabetes subtypes, particularly with longer dosing in the expansion phase of COVALENT-111. The populations that we will be discussing as we review the efficacy and safety data are the modified intent-to-treat population, or MITT, defined as all randomized participants who took at least 1 dose of study drug. This population included patients on active drug and placebo is comprised of 225 patients. The efficacy analysis focuses on those participants that completed the dosing period prior to the clinical hold and were on at least 1 anti-hyperglycemic agent at baseline. This population is comprised of 168 participants, 113 who received icovamenib and 55 who took placebo. The majority, approximately 70% were treated with 1 agent at baseline, primarily metformin, approximately 15% with 2 agents and 5% with triple therapy or 3 agents at baseline. Here, we see the baseline demographics and characteristics. Importantly, they were well balanced between the active and placebo arms. On average, patients were around 55 years old with a mean duration since diabetes diagnosis of just over 4 years. Approximately 40% were female and 60% male with good representation of Hispanics around 50% or 60%. The majority of participants were white and black patients were well represented with 25% of active patients and 16% of placebo-treated patients in this demographic. The mean baseline hemoglobin A1c was just over 8%, 8.2% in active patients and 8.3% in the placebo group, and the mean BMI was around 32 kilograms per meter squared with 60% to 70% of patients being obese as defined by a BMI of 30 or greater. Recall that patients with a BMI lower than 25 were not eligible to participate. Moving now to the efficacy results. Here, we see the placebo-adjusted change in hemoglobin A1c at week 26 for participants suboptimally controlled on one or more anti-hyperglycemic agents at baseline. For all active patients combined, 113, there was a statistically significant 0.36% reduction in hemoglobin A1c at week 26. The left panel looks at the change in A1c in patients treated for 8 weeks, so Arm A, and those treated for 12 weeks, Arm B and C. You can see that the longer duration of treatment resulted in greater A1c lowering with a placebo-corrected reduction of 0.42%, which also was statistically significant. So the longer duration of therapy was associated with a greater decrement in hemoglobin A1c. On the right, we show a comparison of the change in hemoglobin A1c at week 26 for the type 2 diabetes subtypes. Again, in the overall population of 113 patients, a reduction of 0.36%, whereas in the insulin-resistant subtypes, the MARD and the SIRD, a 0.13% reduction and a clinically and statistically significant reduction of 0.73% as we hypothesized in the insulin-deficient MARD and SIDD subtypes. So we now specifically focused on the longer 12-week treatment duration, Arms B and C as well as the insulin-deficient participants, the MARD and the SIDD subtypes, we see that overall Arms B and C combined performed slightly better than the full per protocol population with a 0.42% reduction in A1c. If we then look at the MARD and SIDD patients in the 3 study arms, they had a significant 0.73% reduction in A1c. And when we focus on these patients that dosed for 12 weeks in Arms B and C combined, we see a significant 0.84% mean reduction in hemoglobin A1c at week 26. If we look at only the severe insulin-deficient patients, overall in Arms A, B and C, we see a 0.9% reduction in A1c and with 12 weeks of dosing in Arms B and C combined, a clinically and statistically significant reduction of approximately 1.2%. So these data indicate that 12 weeks of dosing, particularly in insulin-deficient patients results in a robust improvement in glycemic control at week 26. Taking a closer look now at Arm B, which was 12 weeks of dosing and 100 milligrams once daily, we can see that in this arm, particularly in patients in the insulin-deficient subtypes, there were clinically and statistically significant reductions in A1c. Overall, Arm B demonstrated a 0.5% reduction in A1c at week 26. But you can see in the far right of the bar graph that the greatest improvement in A1c occurred in the insulin-deficient patients, particularly the severe insulin-deficient diabetes or SIDD patients, who experienced a 1.5% mean reduction in hemoglobin A1c. So 100 milligrams of icovamenib for 12 weeks in insulin-deficient patients showed the strongest effect across all groups. Here, we see a summary of 12 weeks of dosing, Arms B and C on the top and Arm B alone below, looking at all patients in the per protocol population in these arms and the SIDD and MARD subtypes combined as well as the SIDD subtype alone. For both the combined 12 weeks Arm B and C and for Arm B alone, we can see statistically significant reductions in A1c for all participants in those arms and greater improvements in A1c in the insulin-deficient patients, both the SIDD and MARD patients combined and the SIDD patients alone with the most robust improvement seen with 12 weeks of 100 milligrams daily icovamenib, Arm B in the severe insulin-deficient patients. Lastly, with respect to efficacy, I want to share findings in patients treated at baseline with GLP-1 based therapies. These participants all fell into the mild obesity-related diabetes, or MOD category, which is an insulin-resistant category. Despite this, we saw numerically important mean improvements in hemoglobin A1c despite relatively small number of patients. The combination of icovamenib and GLP-1 based therapies will be an important area of future clinical focus. Turning now to safety and tolerability. Icovamenib was very well tolerated. Here, we see the incidence of treatment-emergent adverse events was similar in icovamenib and placebo-treated participants with 30% of icovamenib patients and 32% of placebo-treated patients experiencing an adverse event through week 26. Importantly, there were no serious adverse events reported, no treatment discontinuations nor study discontinuation due to adverse events, and there were no deaths in icovamenib or placebo-treated patients. Here, we see adverse events that were reported by 5% or more participants in the icovamenib or placebo arms. We also show adverse events reported due to elevations in aminotransferases, ALT and AST, despite the fact that they were reported in less than 5% of participants. The vast majority of these events were Grade 1, so mild in severity. And as mentioned in the previous slide, none led to study medication discontinuation. Importantly, and not shown in this table, there was no hypoglycemia or low blood glucose reported in any participant. To summarize and conclude, icovamenib met the primary endpoint, displaying a clinically meaningful and statistically significant placebo-adjusted reduction in hemoglobin A1c in the prespecified per protocol patient population. Icovamenib was well tolerated with no study drug discontinuations due to adverse events. It displayed the most robust statistically significant mean hemoglobin A1c reduction in patients dosed for 12 weeks and in the type 2 diabetes subtypes characterized by insulin deficiency. In Arms B and C combined, SIDD subtype patients experienced a mean hemoglobin A1c reduction of 1.17%. And in Arm B alone, patients in the SIDD subtype had a mean A1c reduction of 1.47%, both clinically and statistically significant. Icovamenib also demonstrated a clinically meaningful reduction in hemoglobin A1c in the type 2 diabetes subtype characterized by insulin resistance, mild obesity-related diabetes in study participants uncontrolled on a GLP-1-based therapy at baseline. Before I hand over to my colleague, Dr. Alex Abitbol, I want to provide a few words on the clinical hold, which was based on the escalation phase of the study where we had initiated dosing with 200 and 400 milligrams and unexpectedly saw elevations in aminotransferases, importantly, with no elevations in bilirubin. FDA has since provided guidance on maintenance and oversight for patients and lifted the clinical hold. The topline expansion phase results disclosed today show how the current protocol and oversight has worked and has led to a favorable safety profile. Of note, the FDA clinical hold did have a more significant impact in our type 1 diabetes study, COVALENT-112, in which we were not able to complete the planned dosing in over 90% of the targeted patient population. We are, therefore, planning to continue enrollment in COVALENT-112, so we could provide a more complete update with robust data in this patient population in 2025. I will now turn the call over to Dr. Alex Abitbol to provide his thoughts about the clinical relevance of these topline data. Alex?

Thank you, Juan. Good morning, everyone, on this call. My name is Dr. Alex Abitbol. I'm an endocrinologist practicing for about 10 years at LMC Healthcare, Canada's largest specialty care provider in diabetes and endocrinology with 11 clinics and over 20 clinical research sites. We provide care to over 100,000 active patients across Canada annually. I have been a principal investigator on the COVALENT-111 study and contributed to the study enrollment at our clinical research sites. Actually, one of my icovamenib type 2 diabetes patients was presented earlier this year at ATTD Europe in March, where we had shown that the benefits of just 4 weeks of icovamenib's treatment continued even past the observation period. My patient dropped from A1c of 9.6% to 8.9%, just with 4 weeks of icovamenib during the 26-week escalation trial. As I was the treating physician, I continue to oversee the progress and reported that by week 26, the patient had already achieved target A1c levels of 7%, and I was eventually able to reduce and even stop the metformin following the study. Such results are quite impressive, and I have to admit that I have not used an agent capable of providing such robust clinical benefits with nonchronic dosing. In my day-to-day clinical practice, I see many different patients affected with diabetes and generally have access to numerous classes of therapies. The approved therapies all require chronic administration to achieve the desired effect, which is one of the biggest challenges in caring for patients affected with type 2 diabetes. Medication adherence remains a challenge and important consideration today since so many remain concerned about potential adverse effects, unwanted injections and the rising costs of long-term medications. We have now dosed over a dozen patients with icovamenib and have not observed any clinically significant treatment-emergent adverse effects to date. The American Diabetes Association emphasizes early and intensive glycemic control to reduce the risk of long-term diabetes-related complications. This involves adding second and often third-line agents such as GLP-1 receptor agonists or SGLT2 inhibitors to background metformin when glycemic targets are not met. The potential of addressing the pathophysiology of diabetes through a new pathway is very exciting for endocrinologists and specialists alikes, as we need better therapies to address a worsening global epidemic. Icovamenib's mechanism of action, the potential to regenerate beta cells is unique, unlike any other agent I have used. By inhibiting menin covalently, icovamenib is removing the gatekeeper, which is believed to prevent beta cells from regenerating. With the use of icovamenib, the pool of beta cells may then be reestablished, thus, in theory, allowing a patient affected with diabetes to produce insulin independently with their own regenerated cells. I'd like to spend a few moments hypothesizing why we believe we may have observed these very impressive results early on in the trial. The patient of mine that did so well on icovamenib was likely a patient with SIDD, or severe insulin-deficient diabetes. And as we learned today, patients with SIDD seem to be the best responders to icovamenib. It is important to note that patients with SIDD have reportedly the highest risk of major adverse cardiovascular events and cardiovascular death and a more rapid progression to insulin dependence among all subtypes of type 2 diabetes. The patients with SIDD and the MARD subtypes also have an increased mortality risk compared to other subtypes. Patients with SIDD are also the ones who are most in need of an effective therapy that specifically targets the regeneration of a depleted pool of beta cells. None of the anti-hyperglycemic agents used today are capable of addressing the progressive loss of beta cell function. In clinical practice, beta cell decline is fundamental to the ongoing necessity to intensify anti-hyperglycemic therapy in type 2 diabetes to achieve and maintain adequate glycemic control and thus prevent long-term complications. We are still not doing a good enough job as nearly half of patients affected with diabetes do not have well-controlled hemoglobin A1c levels below 7%. I hope that the icovamenib clinical trial program continue to progress and will someday provide us a new pathway to benefit more patients stay better controlled. I am grateful for all of the effort and investment from companies like Biomea and to you, the investors on this call, helping us to reach our mutual goal of better therapies in diabetes.

[Operator Instructions] And our first question will come from Eric Joseph with JPMorgan.

I guess a couple from us. First, on just the treatment completion rate, it seems like it's about 70%. I guess apart from any potential impact from the clinical hold, can you talk about sort of factors relating to compliance rate and whether there is any pattern to where patients kind of may have fell off from participating in the study? And then secondly, the data today on A1c, I think, are all at the 26-week time point, but presumably, you're gathering data sort of over -- on a serial basis. Is the -- can you talk a little bit about sort of the dynamics in A1c decline and whether you're at 21 -- sorry, 26 weeks, you're still somewhat in the dynamic range? Or I mean to say is, does this seem like the nadir of A1c decline or patients somewhat in recovery and how long? Can you talk about the time frame and how you're going to be monitoring A1c going forward?

Eric, this is Juan. I'll take the second question first. I mean we will be showing that, and we'll be presenting the data at an upcoming congress. But it appeared that the A1c had not plateaued at week 26. So when we look, we had A1cs throughout the 26-week period. So we're following these patients, as you mentioned, and we've talked about out to week 52. But that's what we've seen to date. So we're certainly hopeful that we will get even further improvement in A1c over time, but remains to be seen.

That was a compliance question.

Yes. So with respect to patients who did not complete the study treatment, really, it was almost entirely due to the clinical hold. And certainly, from an adverse event perspective, as I mentioned in the safety tables, no patients discontinued study drug or discontinued from participation in the study due to adverse events. So really, any stopping of the medication prematurely, if you will, or the vast majority were due to the clinical hold.

Maybe one last follow-up, if I could. Any -- can you talk about monitoring of C-peptide and -- or HOMA-IR, HOMA-B in the study? And any trends there that may correlate with the pattern and the strength in A1c decline that you're seeing and...

Yes. So we're still -- we did measure it. So during the study, we had a number of oral glucose tolerance tests that we see peptide and insulin was measured. So we will have those data. We do not have them yet though ready to present and validate it. But we will be certainly disclosing that in time.

And our next question will come from Yigal Nochomovitz with Citigroup.

I have a question for the clinical investigator. So in your practice, when you see patients, I mean, do you typically go through this exercise of... [Technical Difficulty]

I think that question was for me, and I'm not sure if it was just on my end, but you kind of cut out during your question. Would you mind repeating it?

I think we lost him.

[Operator Instructions] And our next question will come from Matthew Biegler with Oppenheimer.

Thanks for hosting the event this morning. It sounds like the intent going forward is to focus on the insulin-deficient population. Is that something kind of easy to screen for vis-a-vis insulin resistant? Is it -- what's kind of the FDA's view on using prespecified biomarkers to screen this patient population? And have you had any discussion with the FDA on -- I don't know if there's been an end of Phase II meeting yet, but when you do have it, kind of what's the plan going forward for designing a pivotal trial?

Yes. Matt, this is Tom. Great question. Yes, so we just got the data. So obviously, we're continuing to crank through this data set, and we'll be preparing a package for the FDA. But what I can tell you is that the fact that it works the best in the key mechanism of action, patients, the insulin deficient is exactly what we were hoping for. And an important thing, and I think this is where Yigal was going is you don't have to biomarker these patients in order to develop icovamenib for these patients, meaning in other countries like Japan, they have BMI cutoffs and the BMI cutoffs with their baseline A1c essentially enriches this patient population. So it will be very easy for us to go into the next study and focus with just the MARD and SIDD patient population. And again, it's just looking at their BMI, which controls almost 90% of the biomarker and then baseline A1c. So this is something that's already done around the world, and we're excited to be able to do it for our development plan.

Okay. Got it. That makes sense. I might squeeze in one on COVALENT-112, I guess, because using lessons learned from today, i.e., better responses in insulin deficiency. And I believe type 1 is more of an insulin deficient versus resistant type of problem. So do you expect icovamenib to work better in COVALENT-112? And then have you guided on when we would get an update from -- or maybe top line from that trial? I'd realize 90% of patients had -- were affected by the clinical hold.

Yes. And great question. We absolutely are excited about what we're seeing with 112. It's just that the end is so small. And we don't want to just report on 3 or 4 patients because that's really how many finished the treatment period. And so we think it's much more prudent just to restart the dosing, generate a larger end so that we can have nice confidence in the data. But I can tell you the data we've seen today for type 1 is paralleling what we're seeing with 111 in type 2.

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

This is Lander on for Joe. So first of all, going back to the type 2 diabetes subtypes, what's the HbA1c level threshold that you used to define the SIDD group?

Yes, great question. So there's no specific A1c cutoff. There's 5 criteria that go into identifying each cluster. And we have in our slides, those key characteristics that go into determining the cluster. And so it's not just A1c by itself. There's other factors that go into determine. But essentially, what it's trying to do is assess when this patient comes in the door and gets their A1c measured, how much C-peptide production does this individual have? And that helps them assess, is this patient insulin deficient or really insulin resistant. Insulin-resistant patient population in the U.S., which is about 45%, they make plenty of C-peptide. C-peptide production, although it's adequate from a normal person's perspective, they have so much visceral fat and insulin resistance that they're not able to have adequate glycemic control. So we certainly see effects of icovamenib producing more insulin for these patients. But certainly, it's the deficient that are the most sensitive, which makes sense.

Yes. If I could add on -- this is Juan. I mean for anyone to develop diabetes, they have to have either absolute or relative insulin deficiency. Most people who are overweight and obese and very insulin resistant do not develop diabetes because they have robust enough beta cells to become hyperinsulinemic and compensate for the peripheral insulin resistance. But these subgroups really get at what is the primary pathophysiologic defect, which is leading to hyperglycemia. Is it more of an insulin deficiency state in the SIDD and MARD or more of an insulin resistant state, which is generally associated with extreme obesity. I would say, I mean, I was pleasantly surprised at the proportion of patients that ended up in the insulin-deficient categories in our study despite the fact that our BMI cutoff at the lower end was 25. So we didn't bring in any normal weight patients with type 2 diabetes, which certainly exists out there. So -- which to me shows how prevalent this subtype is even in the U.S. and Canada, where there's a lot of overweight and obesity.

Okay. And finally, so on top of the HbA1c level reductions, do you have data on the number of patients with higher than 1% reductions in HbA1c in each group? Are we expecting to see this data maybe at the 52-week readout?

Yes. We will definitely see it at the 52-week readout, and we'll have certainly more data as we go through this huge data set. So today, just providing the top line results, but we'll have categorical variables as well as time goes on now.

And our next question will come from George Farmer with Scotiabank.

I was wondering if you could comment on whether you observed any weight loss in this trial, number one. And whether any of these patients reached their goal of less than 7% HbA1c, particularly in the SIDD and MARD patient populations? And I have a follow-up.

George, this is Juan. Those are data that we're still looking at and we will be reporting it, but don't have it all fully validated and all yet. So -- but definitely, we'll look at weight. And again, as I mentioned for the last question, the categorical variables also, so percent of patients to target.

Okay. And then -- sorry, Juan, yes.

No.

No. Okay. Great. And then I was really struck by the low degree of liver enzyme elevations. It really just looks like kind of background noise, if you will. Can you describe like how you're managing any sort of liver tox? And it feels like maybe this clinical hold was a bit of an overreaction, at least from my perspective. I was wondering if you could comment on that.

Yes. I'll just quickly comment. This is Tom, before I pass it on to Juan. I think what's important is when we were executing this study, we didn't have the study unblinded, right? So we only had blinded data. And then you don't get to see the full picture what's background, what's not. And I think we're very pleased about the execution of the study from an LFT perspective. And it really comes down to -- actually, these patients finished the treatment prior to the hold, right? So these elevations are clearly not from any modifications to the protocol that were implemented post the hold. And Juan?

Yes. But I would say it also speaks to the dosing regimen that we used. I mean the elevations primarily that we saw in the escalation portion, so the multiple ascending dose portion were in the participants who initiated with 200 milligrams or 400 milligrams. And here, each arm was either exclusively on 100 milligrams through the full dosing period or an Arm C was on 100 milligrams for 8 weeks and then escalated to the 100 milligrams twice daily or 200 milligrams daily in the final 4 weeks, and that certainly may play a role.

Okay. Great. And then one more, if I may. There are some patients out there that had entered the study maybe longer than a year ago. Do you have any updates at least on safety and whether anything suspicious may have cropped up over that time?

No.

We have not seen anything.

Nothing to report.

For my patients, especially since I continue to follow them, no, there haven't been any surprises.

And the next question will come from Yigal Nochomovitz with Citigroup.

Sorry about earlier. But Tom, I think you interpreted my question correctly. But to follow up on that. So for the Phase III, you mentioned just doing it with the BMI cutoff, which simplifies things. Are you also going to use this tool, this clustering tool, the diabetes DDZ tool in doing the enrollment? And then for Dr. Abitbol, I'm just curious how familiar do you believe the sort of general diabetes physician population is with this type of clustering method and whether they're familiar with all these subtypes?

Yes. So I'll take the first one, Yigal. Yes. So from a clinical trial execution, we would just have baseline characteristics, as I mentioned, baseline A1c and BMI cutoff to enrich to the MARD and SIDD patient population. So there's nothing that we need to do outside of just setting the inclusion criteria.

And from a clinical practice setting, well, you touched on a good point. I don't think that there's necessarily widespread lay -- or sorry, day-to-day clinical acceptance for these subtypes today. That being said, this is certainly an emerging field, particularly in type 2 diabetes. And Juan already touched on sort of the absolute deficiency end of things in type 1 and more of the classic insulin resistance in type 2, but there are so many patients that fall into a lean body mass index category, a younger age, older age that we tend to be treating them differently in clinical practice as you wouldn't necessarily use robust agents like Ozempic or Mounjaro in somebody who's fairly lean, although some of us do. And so there will be more data, more research, more descriptive and clinical basis for sort of differentiating these subtypes in years to come, both for this and other therapies.

And our next question will come from Tony Butler with Rodman & Renshaw.

Dr. Abitbol, you've made a reference to a patient who you had dropped their metformin use because they had reached appropriate A1c. So the question becomes, how do you know when to drop someone from their existing therapy, were that to be a goal? And number 2, is a goal to actually drop individuals from their previous therapy, especially if they're on 2 or 3, which could be very inconvenient to something that's certainly much more manageable for the patient? And how would you know when to do that?

Absolutely. So it's a fantastic question. I think that nowadays, that question has probably changed a little bit. From a safety standpoint, if somebody is achieving target A1c, both less than 7% or less than 6.5% and is maybe now even approaching A1c levels in the normal range of less than 6% or less than 5.7% depending on what country you practice in, well, if they're on background therapy that can cause low blood sugar like insulin or older agents, we're implored to stop therapy because it's more harmful than to keep A1c that low. Further to that, there's GLP-1s and SGLT2 inhibitors that in the right clinical context should probably never be held because they provide cardiovascular and renal benefits in a specific population of patients that has demonstrated this benefit. But certainly, in clinical practice, I don't mean we're deescalating therapy when you achieve a target. This is when you're well below the target. When somebody who's obviously made changes seen an incredible response to therapy has possibly seen reductions in weight and now we slowly reduce some of the agents that don't provide any of the cardiorenal benefits. You do so with the recognition that you're dropping them in the hopes of not just giving them back that same therapy at the next visit. So it's usually when they both achieved target, maintained it and are, of course, well below the usual glycemic targets. But these are the exceptions. This is not something you see very often. As people get older, they become more sedentary. Diabetes is a progressive disease. So these tends to be more the exceptional patients, and I hope that we'll see more of them like this in the future.

And our next question will come from Srikripa Devarakonda with Truist.

Looking at the data, the combined data from MARD and SIDD, I was just wondering if the SIDD subgroup was driving the efficacy. And also, given that the data is from small cohorts, right, 7 SIDD patients, especially with the particular dose that you're looking at and combined with 13 patients, what's your level of confidence in the magnitude? I mean you're seeing benefit for sure, but the magnitude of benefit that you're seeing? And then maybe for Dr. Abitbol, based on these data, where and how do you think icovamenib would be used, especially in the context of GLP-1 agonist-based therapies?

Yes. This is Tom. Great question. So we're actually very confident, and we're confident because of just the consistency of the impact that we're seeing within the beta cell deficient patient population just as a whole, if you look at SIDD and MARD combined. And certainly, MARD is not as beta cell deficient or insulin deficient as SIDD. And so looking at MARD and SIDD combined versus SIDD, it makes sense why SIDD would have then a deeper A1c reduction. They certainly make less C-peptide. And so this just really matches well with the mechanism of action.

And with regards to your question as to where do I see it fitting in along the treatment landscape. So you touched on obviously why this particular molecule has at least impressed me early on in its clinical development because it's really any entry point along the pathway of either developing diabetes, having prediabetes, having diabetes on one agent being on multiple daily injections of insulin. Well, because it essentially affects the pancreatic beta cell pathway, there's really any entry point you might consider it because it will either benefit the patient in terms of A1c reduction, allow a diminishing, slowing down or reversal of the progressive nature of diabetes as even patients who are on GLP-1 receptor agonist often still see progression and ultimately do require insulin down the road. But in my opinion, I think the lowest hanging fruit for this particular molecule of cost coverage, reimbursement and access really are all equal would be even before they develop the diabetes to identify people at risk of developing the disease and catching them early, thus limiting any potential harm that glucose will do to this person in the future.

Great. And if I can ask a follow-up question to the company. From a dosing perspective, you talked a little bit about how -- what you're seeing between the 12 week and the 26 week. It's not plateaued yet at 26 weeks. But from a dosing and regulatory perspective, do you have a sense of how frequently you may have to administer the drug? And how do you incorporate that into a trial?

Sorry, Srikripa, I missed the last point or the last piece of the question. How quickly we need to dose it? I don't know what you mean?

No, not how quickly. So the dosing and how frequently you may have to administer the drug?

Yes, how long will the 12-week therapy last for the durability...

Yes. Exactly. And how do you incorporate that into a clinical trial because it could be very long, right?

Yes. So as we mentioned in the beginning of this year, the FDA made it very clear and aligned completely with our clinical thinking is that for this molecule, week 26 is the primary endpoint. So for late-stage development, that would stay true that week 26 is the primary endpoint and then week 52 is the key secondary. So that will be the design of the trials going forward. And so obviously, we want to see how week 26 performs and then how -- what does week 52 look like, what does that durability look like. In terms of how frequently this drug needs to be dosed, that is still unknown. And I think that will come about as we continue our development.

And it may vary from patient to patient as well. And it is something that could be incorporated into the future clinical trial as well.

And our next question will come from Peter Lawson with Barclays.

Great. Thanks for the data and thanks for the update. Really appreciate it. Just on the back of Srikripa's question around the patient numbers in the -- in Slide 15. How dispersed were the response rates? I just didn't see error bars. I know you've got p-values, which is great, but just curious on kind of if they're kind of single patients that are kind of driving that response or if it's kind of a tight reduction in A1c?

Yes. Great question, Peter. For the MARDs and the SIDDs, it's very tight. In terms of the response rate, it's quite high. Those data, obviously, we'll share at major medical conferences. But the reason why you're able to get a p-value on such a low end is because it's so tight.

Got you. Okay. And then the MARD patients, I didn't see those broken out and I didn't see Arm C broken out. How do those patients look?

Yes. Great question. So we didn't break out C. As you can see, C is a very low end when you get to the MARD and the SIDD subgroup. So -- and Slide 16 kind of highlights that. You can see the end of B, C combo in Arm B alone. So C is certainly a very low end by itself. And so that's why we combined B and C together, so you can get a nice feel for what 12 weeks of dosing does in general. And it's just an end thing.

Got you. And then how did the MARD patients by themselves do?

How does the MARDs. Yes, so if you look here on the chart, especially on Slide 16, the MARDs even in all arms are performing really well. And MARD, obviously, is the higher end because we break -- we give you the SIDD. So you can see that MARDs do just -- they do very, very well. They're not as deficient as the SIDDs, but they're doing well.

And our next question will come from Joe Catanzaro with Piper Sandler.

Apologies if these have been asked, I've been jumping around. So can you guys say what the placebo A1c reduction was at week 26? And when we're looking at the subgroup analysis, is the placebo number, all placebo patients or we're looking at various subgroups? I guess my question is whether SIDD placebo-treated patients would have a different A1c reduction at week 26 in the pooled cohort. Hopefully, that makes sense.

Yes, very good question. So overall patients, the placebo change was 0.08 plus. And then when we do the analysis in the MARD, the SIDD/MARD combos, those are not placebo of all patients, but placebo of SIDD and MARD, respectively, in their groups. Because you're right, the placebo do behave differently in different phenotypes. So we've adjusted appropriately so that when you see the placebo adjusted for SIDD and MARD combo or SIDD by itself, it's placebo corrected by that SIDD placebo patient. To make it apples-to-apples.

Okay. Great. And then maybe on the dosing paradigm, I know the trial allowed for retreatment. Has there been any retreatment?

Yes. So I don't know, Juan, do you want to talk about it?

Yes, there were actually relatively few retreatments. So we'll take a look at that, but it's very few retreatments out at week 22. And some of those patients, the clinical hold occurred before patients reached that even potential to retreat. So there's not much there, but it's a few patients that did -- that were retreated, and we'll have those data in time as well.

I show no further questions at this time. I would now like to turn the call back to Thomas Butler, CEO, for the closing remarks.

Thank you, operator. With the data we presented here today, we believe we have achieved proof of concept and validated a new pathway to address diabetes. The data we released today has the potential to help transition from addressing the symptoms of diabetes to treating the root cause, a depleted pool of beta cells. We are very excited about today's data and the potential of icovamenib in type 2 diabetes. Thank you for your interest in Biomea. This concludes today's conference call.

Thank you. This does conclude today's conference call, and thank you for participating, and you may now disconnect.