Madrigal Pharmaceuticals, Inc. (MDGL) Earnings Call Transcript & Summary

Good day, ladies and gentlemen, and welcome to the Madrigal Pharmaceuticals Resmetirom AASLD Clinical Data Review Conference Call. [Operator Instructions] As a reminder, this call may be recorded. I would now like to introduce your host for today's conference, Dr. Paul Friedman, Chairman and Executive Officer of Madrigal. Dr. Friedman, you may begin.

Thanks. Hello, and thanks for joining us on the call today. This is the fine print dealing with forward-looking statements we might make today. Please see our latest SEC filings for a summary of the risk factors associated with our business. The primary purpose of the call is to review the latest data on resmetirom from the open-label portion of our ongoing Phase III MAESTRO-NAFLD-1 study, which was just presented at AASLD. Dr. Harrison will take us through these data, Dr. Stephen Harrison. We're also entering an exciting period for Madrigal with the planned releases in the coming months of top line placebo-controlled double-blind data from each of our 2 Phase III clinical trials, MAESTRO-NAFLD-1 and MAESTRO-NASH, and Dr. Taub -- Rebecca Taub will provide a recap of these studies, what to expect and when. And I should point out that Dr. Harrison is the Medical Director of the Pinnacle Clinical Research Center in San Antonio, Texas, and he's a visiting professor of hepatology at Oxford. Becky is CMO, President of R&D of Madrigal, and a founder of the company. At the end of Becky's remarks, we'll open the call for Q&A. Before turning the things over to Stephen to review the data, I'd like to take the opportunity to provide a few words on Madrigal and on resmetirom. To begin, and importantly, Madrigal has an experienced management team with proven track records in bringing multiple drugs through discovery, development and onto the market. Becky, our CMO, is widely recognized as a leading expert in liver disease and NASH. I've been involved in the successful development of multiple new therapies. And Remy Sukhija, our Chief Commercial Officer, has launched more than a dozen drugs. We're confident we have the scientific, clinical and commercial resources to succeed in bringing resmetirom to market and to emerge as a leader in the treatment of NASH. Now there's an obvious significant need for a safe and effective therapy for NASH. And currently, there are no approved therapies. We believe that resmetirom, a once-daily oral therapy, through its unique transcriptional program, gets at the underlying cause of NASH and it has first-to-market potential. Our Phase II clinical trials demonstrated compelling efficacy and an attractive safety profile for resmetirom. Based on those results, we designed an extensive Phase III clinical development program, which Becky will describe later. We're confident that these studies are appropriately powered and of sufficient duration to determine safety and therapeutic benefit. We're targeting an NDA submission in the second half of 2022. And our plan for the launch is to commercialize resmetirom on our own in the United States and to secure a partner or partners for U.S. I should also mention that 2 other people who are on the call today are our CFO, Alex Howarth; and I mentioned Remy Sukhija, the Chief Commercial Officer. With that, I will turn the call over to Dr. Stephen Harrison.

Thank you, Paul. It's great to be here this morning coming to you from Texas. So just as we wrap up AASLD, it's an exciting time. I think, as Dr. Manal Abdelmalek mentioned in her NAFLD Debrief last night, despite COVID, significant advances in NAFLD have continued throughout this pandemic and resmetirom continues to be a leader in this regard. With that, I would like to highlight some of the new data that we presented and my colleagues presented relative to resmetirom and its developmental program. So this is just an overview slide of all the studies that we have done or continue to work on. And just to highlight again that with building on the Phase II data that we published in Lancet, we have 2 ongoing Phase III trials. The registration trial, MAESTRO-NASH, has completed its subpart H enrollment, continues to enroll towards the Phase IV portion of that. And again, their Phase IV portion is a 54-month outcome liver-related event situation. For MAESTRO-NAFLD-1, over 1,200 patients were enrolled. This is a 52-week safety lipid, NASH biomarker and imaging study. There were an 80-milligram, 100-milligram and placebo arms in that trial. There was also an open-label arm that we'll speak more about today. There's also a NASH cirrhotic open-label arm. And then there was an additional open-label extension, and this is after completion of MAESTRO-NAFLD-1, patients are offered further rollover for additional treatment duration of 52 weeks. These trials provide a comprehensive data set to support the efficacy and safety for accelerated approval of resmetirom for the treatment of patients with NASH and advanced fibrosis. Next? So just to review, highlight at a high level, the mechanism of action and why we think thyroid hormone selective targeting is important in underlying NASH. If you recall, TSH is generated from the pituitary gland, it has actions on the thyroid gland. T4 is generated, that's inactive. It's taken to the liver and convert it to T3, the active component. And we know that in humans, if we modulate thyroid hormone receptor beta, we can lower LDL, triglycerides, liver fat and potentially impact NASH. And it's important to note that resmetirom, which is a thyroid hormone beta selectively targeted molecule, has no THR alpha effect, so therefore, no thyrotoxicosis. It has a proven safety track record and efficacy track record, now dosed in more than 1,000 subjects. There is no exposure in tissues outside the liver. And these pleiotropic effects that we're seeing, we feel really have the potential to hammer away that not only the histopathologic effects of NASH but also some of the extrahepatic complications seen in this disease. Next slide? Just to dive in a little bit more on how the thyroid hormone receptor beta pathway plays a key role in liver health. We know that if you agonize THR beta that it acts on multiple pathways that are thought to be principally indicated in underlying pathogenesis of NAFLD to NASH. This has effects on de novo lipogenesis, fatty acid oxidation, mitochondrial biogenesis, cholesterol metabolism and I might add direct anti-inflammatory and antifibrotic effects modulating through the TGF-beta pathway. Now I think there are 3 principles that are important to note in NASH. We know that most hepatic fat comes from external sources, particularly free fatty acids from adipocytes. And in NASH, beta oxidation is reduced, therefore, allowing free fatty acids, which are toxic to the liver to wreak havoc on hepatocytes. In human NASH, the liver also has relatively low THR beta activity, which worsens mitochondrial dysfunction and lipotoxicity. I think it's also important to add that since we see robust effects on liver fat content reduction with resmetirom, we often get asked, where does the fat go? Well, the fat actually gets burned and the fat gets eliminated through carbon dioxide respiration. And this is well known to the citric acid cycle. Next slide? So here we are. I want to show you a new analysis of the 36-week Phase II NASH patient data really juxtaposed against some of the new NASH cirrhotic open-label treatment, liver volume assessment. And we'll talk a little bit about the noncirrhotic arm as well. Next slide? So just to highlight and again, summarize the results of the Phase II, remember, the primary endpoint was met, which was a relative reduction in hepatic fat at week 12. And there, when you compare the results of resmetirom at week 12 and week 36 to placebo, we see on the left-hand bar graph a significant improvement. And then maybe just as importantly, we see a dose response between 60, 80 and even 100 milligrams such that dose -- there was a dose-dependent 50% reduction of hepatic fat at 80-milligram. Several key secondary and exploratory endpoints were achieved that were listed here, highlighting that there were statistically significant reductions in fibrosis biomarkers as well as LDL cholesterol and other atherogenic lipids, and it was deemed to be well tolerated and safe. Ultimately, in the blue box here, I think the take-home point was that resmetirom responders with at least a 30% PDFF reduction, had higher rates of NASH resolution compared to nonresponders and that has given us a lot to build off of as we move the program forward. Next slide? Just to highlight again, we talk a lot about liver fat content reduction, the effects of resmetirom on NASH. I think it's also important to highlight that while fibrosis wasn't powered to show a benefit with resmetirom, we certainly saw signals that this was happening, not just in the movement in a positive direction in the biomarkers, but also what we noticed was that if you resolve NASH, you also tended to resolve fibrosis, and at least 61% of NASH resolvers had at least a 1-point improvement in fibrosis. And when we look at a more specific quantitative evaluation of collagen by Second Harmonic Generation, which is Histoindex's platform, half of the F3 patients had at least a 1-point improvement in fibrosis compared to no improvement in the placebo group. Next slide? So jumping in here to some of the data presented, this is liver volume assessed in the NASH Phase II and the resmetirom-treated NASH cirrhotic patients from the MAESTRO-NAFLD-1 open-label arm. And at the top, you can see there are 3 columns. There's the Phase II placebo group, the resmetirom group and then the third column is the open-label cirrhotic cohort with 105 patients evaluated. And there, it's important to note that all of the patients receive resmetirom. The key message here is liver volume is markedly elevated in patients with NASH. As you see in the first 2 columns, liver volume of 2,100 and 2,352. And as importantly, it's -- just as importantly to note that in the cirrhotic cohort, we see also significant increases in liver volume. Now comparing that to healthy adults, we see around 1,400 CCs. So around 2,100 to 2,300 is quite a significant increase in the size of the liver. And it's also important to note, at least relative to the third column here, the open-label cirrhotic, that the baseline NASH cirrhotic liver volume was increased more than expected based on a lower MRI-PDFF of 8.1%. Next slide? Now just jumping into this, we're going to look at 3 different graphs. We'll start with the first one on the left, and this is PDFF and liver volume time course, focusing on 4 different lines. So just to walk you through, the first 2 lines in blue, the light blue and dark blue, represent placebo. And when we look at the red in the black, that's the resmetirom treatment. Now on the vertical axis is PDFF percent and on the horizontal axis is time, and way over on the right is liver volume. And it's important to note that when you look at the placebo, liver volume and PDFF, there's very little change. However, when you look at resmetirom, you see a sharp decrease to week 12 in liver volume and in PDFF, and that was significant. Now when you look at the center and we look at PDFF versus liver volume reduction, a couple of interesting things to point out. First looking at the blue bars with placebo and looking at week 12 PDFF percent change from baseline. If you do not move liver fat more than 30%, the liver volume doesn't change. If you move it more than 30%, even in placebo, we're not seeing a huge movement in liver volume. It's down about 3.4%. Now compare that in the red bars, this is now resmetirom, same parameters, week 12 PDFF percent change from baseline. And those that had less than 30% liver fat content reduction, we're still seeing a nice drop in liver volume that's accentuated to 21% if you have at least a 30% relative reduction in liver fat. Now looking at the cirrhotic cohort. Let's first look at the red bar graphs. This is PDFF percent change from baseline. And if we have very little fat at baseline by PDFF, meaning less than 5%, we see about a 4% change in liver volume. However, if that is more than 8% at baseline, the liver fat content is more than 8%, meaning a fatty liver, we are seeing about 32% reduction in liver volume now -- or in PDFF, I mean. Now if you look at the graph on the left, this is the brown bars, it's important to note that the change in liver volume of 15% and 18% is consistent with what we're seeing in the middle bar graph where you have 13% and 21% reduction. So with the cirrhotic cohort regardless of whether there was a lot of fat at baseline or little fat, we're still seeing good improvements and changes in liver volume. So it's a complicated slide, but it's important to note that liver fat reduction in NASH cirrhotic patients was similar to noncirrhotic -- I'm sorry, liver volume reduction in the NASH cirrhotic patients was similar to noncirrhotic NASH patients. And importantly, it was largely independent of the PDFF reduction. Next slide? Now let's shift gears to the MAESTRO Phase III clinical trial and screening parameters. Next slide? So just again to remind you of the MAESTRO-Resmetirom Phase III study design, MAESTRO-NASH ongoing Phase III ,52-week serial liver biopsy trial in patients with NASH and fibrosis. And when you combined both MAESTRO-NASH and MAESTRO-NAFLD-1, these are 52-week Phase III trials. They provide a comprehensive data set in over 2,000 patients to support efficacy and safety. And it was designed to support accelerated approval for the treatment of patients with NASH and significant fibrosis, F2, F3. Both trials importantly, employ noninvasive readouts that allow for a framework for diagnosis and monitoring of patients, treatment response. And looking at the open-label MAESTRO-NAFLD-1, it provides ongoing data readouts designed to support safety and potential benefits of resmetirom. Next slide? So just to speak to the MAESTRO-NASH and MAESTRO-NAFLD-1 screening algorithm. Again, here, looking at -- first of all, we feel like you need to have at least some metabolic risk factors to ideally find patients that have NASH with underlying fibrosis. So we know the risk factors on the left, age, BMI, type 2 diabetes, hypertension, dyslipidemia; and noninvasively, we know that FibroScans over 8.5 kPa link to fibrosis and CAP scores over 280 link to steatosis. So when we look at the 8-week screening process for MAESTRO, we want to look at combinations of these, patients with at least 3 metabolic risk factors, noninvasive tests that meet these thresholds as described in FibroScan. And in medical history, we then have 3 different buckets. We order labs, those labs are consistent with inclusion and exclusion criteria. We then do MRIs, MRI-PDFF, looking for patients with at least 8% liver fat. And then liver biopsy and you see the parameters there focusing on F2 and F3 patients, but we will take F1Bs or F1As and 1Cs, where there is additional data. So F1B is for sure, F1B and 1C with a PRO-C3 of at least 14-nanogram per ml or histologic -- or historical liver biopsy. Now when we do all that, what we found is that 80% of the screened MAESTRO-NASH patients have had NASH with significant fibrosis. So we know this algorithm works. In addition, noninvasively, an MRI, particularly in MR elastography, was obtained in more than half of patients, although this was not used as an eligibility criteria. Next slide? Now looking at this MRI data and other noninvasive data, we put together what's called the MAST Score, MAST. Now this is an algorithm for predicting active NASH with significant fibrosis. And it's based on a combination of the serial marker, AST, MR elastography and MRI-PDFF scores. Importantly, more than 1,000 patients have had these tests done and compared to liver biopsy in the screen population from MAESTRO-NASH. In MAESTRO-NAFLD-1, patients had noninvasive tests also that allowed for the calculation of MAST at baseline and after treatment with resmetirom. So that's an important set of data to look at. And then furthermore, analysis of the baseline MAST compared with liver biopsy showed the following: MAST is predictive of fibrosis stage in NASH and of the level of NASH activity. MAST may rule out early fibrosis stage better than FAST, which is a published algorithm of FibroScan and AST, allowing focus on patients with significant fibrosis. This becomes important when we want to identify patients for potential treatment with resmetirom. Elevated MAST score in the setting of metabolic syndrome may predict NASH with significant fibrosis without a need for liver biopsy. So this is that continued work on moving beyond the biopsy. Next slide? So just to show you some spearman correlations with FAST, again, that's FibroScan and AST, compared to MAST, which is MRE, MRI, PDFF and AST as well as individual component, what we see is on the first 2 columns, we have NASH or NAS, NAFLD activity score, we have the correlation and the p-value. The last 2 columns are fibrosis stage, again, the correlation with its p-value. And we see when you compare FAST to MAST, that actually there is a better correlation for both NAS, accentuated a bit more for fibrosis stage. And then when you look at FibroScan, you see its corresponding correlations for NAS and for fibrosis. It's better for VCTE than it is for CAP. CAP is just a predictor of steatosis. It doesn't really link that well to NAS or fibrosis stage, but we do see where FibroScan does. However, when you look at MR elastography, we see relative to VCTE, a better correlation, particularly for fibrosis stage. And so what we know here from this data set is that MAST and MR elastography had stronger correlations with fibrosis stage 1 biopsy than FAST or FibroScan. Next slide? So now turning to the MAESTRO-NAFLD-1 open-label 100-milligram arm, and more recent data presented at AASLD. Again, this is the study designed to remind you of the MAESTRO-NAFLD-1 study. It is a randomized, double-blind, placebo-controlled trial. There is an 80-milligram and 100-milligram blind -- and placebo arm, those are blinded, but there is an additional 100-milligram open-label arm, and that's what we're going to focus on here. There were 171 patients included in that open-label arm and there were noninvasive measures to assess safety and efficacy that we'll speak about. Sometimes there's a bit of confusion here, and I want to clarify that there is an additional open-label treatment of special populations, that being renally-impaired patients and those with compensated cirrhosis and there are over 100 patients in the compensated cirrhotic open-label arm as well. Next slide? So looking at the non-cirrhotic open-label arm. This was again an exploratory evaluation of safety, imaging and biomarkers. And a total of 171 patients were enrolled in this open-label 100-milligram, 52-week arm. And there were a number of 52-week measurements obtained to include lab tests, safety analyses, MRI-PDFF, MR elastography and FibroScan, and all were included in the safety population. 94% and 89% completed key efficacy endpoints at weeks 24 and 52, respectively. The primary endpoint of MAESTRO-NAFLD-1 was safety. However, there were key efficacy objectives, and we'll show some of those today. Next slide? So looking at baseline characteristics of these 171 patients enrolled in this open-label non-cirrhotic arm, there's a lot here. I'm going to highlight the important variables on the right. These included key demographics, mean age 55, majority female, they were significantly overweight, BMIs of 36, about half were diabetic, 70% or so were hypertensive, more than 70% had dyslipidemia, about 44% were hypothyroid. And I think it's important to note also that the mean ASCVD risk score was 11.6%. Now looking at noninvasive test FibroScan mean 7.7, PDFF 17.8, and comparatively, MAESTRO-NASH baseline FibroScan mean is higher and consistent with F3 NASH. Next slide? Now I'm going to break this apart. It's a busy slide, but it's important, I think, to go through this. When we look now first, at the bar graphs, these are week 52 data. And on the left, we see MRI-PDFF. The percent change from baseline is 53%. When we look at liver volume corrected change from baseline, it increases to 61%. And even with CAP from FibroScan, the controlled attenuation parameter, we're seeing about a 53% change. Looking in the middle bars at MR elastography and FibroScan KPA, we see a change of about 0.4, which is significant. We have a baseline MRE at 3.5 and dropping by about 0.4 with MR elastography. With VCTE, we see about a 2-point-change in KPA. When we look at the last three -- last set of bar graphs, where we look at MR elastography and VCTE, now we see a percent change from baseline of around 10% for MRE and 20% for VCTE. Now -- when overall, what does that mean? Well, approximately 50% of patients had a 15% reduction in MR elastography and/or a 20% (sic) [ 25% ] reduction in FibroScan kPa. Historically, when we put that in context, what we know in the data is that worsening of FibroScan KPA by 20% and MRE KPA by 15% is associated with disease progression. And so I think that's important context to have. Now looking below at serum markers, adiponectin, reverse T3, CK-18 and ELF baseline to week 36, 52, we see that these biomarkers showed reductions over the time course of the study that were significant across the board. Next slide? Now looking at MAST, we talked about what that meant in the previous slides. Now let's apply it to the MAESTRO-NAFLD-1. In the first set of bar graphs on the left, we see the percent change from baseline in MAST. First, looking at all-comers, week 16 to week 52, we see continued reduction percent change from baseline of 26 at week 16, and 48 at week 52 that reach significance. And then when we focus on those with higher MAST scores, baseline values greater than 0.12, we see that week 16 and week 52 changes from baseline are accentuated even more. And then on the right, we can look at the MAST score time course. And the red bars are all-comers, baseline week 16, week 52. And we see that if you focus on those with a higher baseline MAST score, over 0.12, you see an accentuated response over time. So in conclusion from this graph in patients with baseline MAST greater than 0.12, which links to a median baseline MRE of 3.4, the MAST score was reduced at week 16 by 50%, and at week 52 by 63%, with a reduction of about 0.134. Next slide? Now here, just to show you some MRI-PDFF and liver volume reduction across the subgroups. I just want to highlight a couple of things here. If you look at sex hormone binding globulin that's low, we don't tend to move on a least square mean value PDFF as much as if you have a high SHBG. And again, that's a marker of target engagement. And we know that if you lose more than 5% weight loss, that's going to have a bigger impact on PDFF. Same thing we see with liver volume on the right, low SHBG. We're not moving it as much as if you have high SHBG at week 52 and same thing for weight. Overall, liver volume is elevated in patients with NASH with NAFLD by about a 30% to 50% margin compared to controls. And with resmetirom, there is reduced liver volume at week 16 and week 52 of 21% and 23%. In all demographic groups, liver volume reduction was, on average, about 450 CCs. And importantly, liver volume reduction was greater than predicted by percent reduction in PDFF. The average liver volume corrected PDFF reduction at week 52 was 61%. Next slide. Now again, just continuing to highlight improvements in liver chemistry tests, we see over the time course that ALT, AST, and gamma GT, all continue to improve, as you see on the left here in a line graph. And it's important also to note that these change from baselines are quite significant. ALT values that were at least 30 units per liter or higher at baseline had a 20 unit per drop over the time course, which is highly significant. Next slide? So what about atherogenic lipids? Well, resmetirom reduced markers of CV risk. We know CV disease is increased in NASH patients, and we see reduced LDL-cholesterol, ApoB, triglycerides and lipo(a), which are key secondary endpoints in the MAESTRO studies. Furthermore, there is a decrease in blood pressure, which may reflect improvement in metabolic syndrome. There's a small overall net weight loss, 21% lost, at least 5% body weight, 9% increase body weight by at least 5%. Next slide? So in summary, resmetirom at 100-milligram per day was well tolerated. I think that's an important point to drive home. There's a high study completion rate, a low adverse event-related dropout of only around 1.2%. Consistent with published data, generally mild diarrhea or increased stool frequency at the beginning of therapy, incurred in about 10%. It didn't lead to study discontinuation. Important also, no central thyroid access changes or adverse effects on vital signs and no change in bone mineral density at week 52. Next? So in conclusion, in this 52-week Phase III open-label study of resmetirom, once-a-day oral medication, non-invasively identified NASH patients that were treated with 100 milligrams of resmetirom demonstrated significant improvements in hepatic fat liver volume. Fibrosis as assessed noninvasively by ELF, MRE and FibroScan, we see improvement in liver cell injury and inflammatory biomarkers as well as LDL and atherogenic lipids. We know it's well tolerated. There are some limitations of the study that include relatively early patient population from a fibrosis perspective in the absence of a placebo control, but this study does highlight the potential use of NITs to diagnose NASH, and importantly, monitor individual NASH patient response to resmetirom. With that, I'd like to turn -- turn this over -- actually, finally, let me just make one more conclusion, and then I'll turn it over to Dr. Taub for her remarks. It's important to note that targeting the hepatic thyroid hormone receptor pathway is promising and is a differentiated strategy to treat NASH for many of the mechanisms that are in development currently. Resmetirom decreases lipotoxicity and reverses hepatic hypothyroid disease that's present in NASH patients, and this allows for restoration of normal liver function. New today, resmetirom reduces liver volume which is greatly elevated in both cirrhotic and non-cirrhotic patients. And also, it's important to note that resmetirom-mediated liver volume reduction is greater than predicted from just PDFF production. In the Phase II trial at lower doses of 60 and 80 milligram, we saw a reduction in liver fat and NASH endpoints. Liver fat and liver volume reductions were associated with improvement in all NASH biopsy components, including fibrosis stage. And finally, the Phase III MAESTRO-NAFLD-1 100-milligram dose was well tolerated and demonstrated multiple NASH biomarker and imaging improvements. Now I'll turn it over to Dr. Taub for her remarks. Thank you.

Thank you, and hello, everyone. So I'm going to just briefly go over expectations for development timing and data readouts, showing you last year, this year and our prospective readouts for next year. So by the end of this year, we're expecting initial top line data from the blinded arms of MAESTRO-NAFLD. This study completed the enrollment in October of last year. It's a 52-week study. And the -- completing the dosing occurred in October last month. However, the patients are actually completing the trial this month by the end of November. So we will be continuing to wrap up the data from that study, clean the data, as you know, this typically takes a couple of months from end of a clinical trial. However, our plans are that we will be able to release the top line efficacy endpoint from -- particularly from the earlier time points, the 24-week lipid endpoints and the 16-week MRI-PDFF by the end of the year. And then in Q1 of 2022, we will be able to following database lock for the safety database, provide additional data from the MAESTRO-NAFLD study. And then we will also be providing additional data that you've seen some of this data from the MAESTRO-NASH cirrhosis population in the upcoming months. And then critical time line is for Q3 of 2022 when we will have a top line liver biopsy data from the pivotal serial liver biopsy trial on MAESTRO-NASH. So with these brief comments on our development time lines, we'll turn it over for Q&A.

Yes. Operator, could you open things up for Q&A, please?

[Operator Instructions] And our first question comes from Andrea Tan from Goldman Sachs.

I just had one, based off of a meet-the-expert session this weekend, where some of the participants debated the use of biomarkers and liver biopsies and clinical trials. Paul or Becky, just based on your conversations, how groundbreaking will the findings in MAESTRO-NASH and NAFLD-1 be to changing the thinking here? And how open do you think the agency will be to incorporating these surrogate endpoints on the forward?

Yes. So I would also like to say that Dr. Harrison has a lot of insight into the various ways of diagnosing NASH and the biomarkers and imaging that are used to follow a patient response. So we'll also let Stephen add his comments here. But just to say that one of the goals of our company all along has been to really do very extensive biomarker and imaging assessments that will help us to diagnose NASH and then be able to follow treatment response to resmetirom. And I think that you see from the MAST data we provided that we have a humongous data set of FibroScans, MREs, MRI-PDFF and lots of different biomarkers. And that our success rate after identifying patients with metabolic syndrome and screening for MAESTRO studies, was very high in terms of identifying patients with confirmed NASH with significant liver fibrosis on liver biopsy. And so I'll turn it over to Stephen for further comments.

Yes, I think that's -- thank you, Becky. I think that's a great comment, a great question, great response as well. And I'll just say the Liver Meeting, there was a tremendous amount of data on noninvasive tests that were presented. And I think Alina Allen from the Mayo Clinic did a great job of summarizing what we know about NITs relative to NASH. And remember, you have to look at this through 3 different contexts of use. Number one, diagnosing the patient that needs to be treated. Number two, measuring therapeutic efficacy, do we have tests that can do that? And then finally, long-term outcome data. And as I mentioned in my post-grad course talk, we have quite a bit of data building on long-term outcomes with MR elastography. And one of the nice things about MAESTRO-NASH and MAESTRO-NAFLD-1, is we have a large volume of MRE data being collected. One of the things also that Dr. Allen mentioned in her review of these biomarkers, was it's important at the primary care level to have a test with high negative predictive value because we know that the majority of fatty liver patients do not have underlying fibrosis or significant fibrosis. So we want at a primary care level, provide a test that can exclude disease. And then at the referral level, have a test that has high positive predictive value to rule in those patients that need to be treated. And there, we usually will look at combinations of therapy or combinations of treatment. And that's where MAST comes in, that MRE, MRI-PDFF and AST that we brought forward at this liver meeting. Now it's not the only one. There are others out there. Professor Rohit Loomba has looked at method, which is looking at MRE and . But the point is, at the primary care level, we need a test to rule out disease at the referral level a test to identify those that need to be treated more appropriately. And when you look at the MAESTRO-NAFLD-1 and MAESTRO-NASH programs, it really is collecting a large body of NITs that will allow us to begin to answer all 3 of these context of use.

And our next question comes from Liisa Bayko from Evercore ISI.

I just sit here and reflect on the data collectively. Can you maybe comment on some of the read-throughs you see here in terms of some of the parameters to what effects you might see on changes in NASH and fibrosis? Because we know a lot of these things are kind of biomarkers that relate to those things and can be predictive. So what kind of changes do you think that this is really reflected?

So I'll comment and then allow Stephen to comment if he has any additional comments on this. But one of the issues with the fibrosis endpoint on liver biopsy is that really more due to sampling, I would say, than disagreement between pathologists, that there is some variability in fibrosis stage that's detected on liver biopsy. And it's a very categorical measurement. So in order for the fibrosis endpoint to be observed, you do need a large well-powered study, which is what -- exactly what we have with MAESTRO-NASH. We observed using a more linear assessment in Phase II, Second Harmonic Generation, that there was a reduction in fibrosis, particularly in the more advanced patients with F3 and that we saw resmetirom impacting in a significant way, multiple fibrosis biomarkers as well as reducing NASH. So we believe that the Phase III study is very well powered for the NASH fibrosis endpoint and that we will achieve it based on the data that we saw in Phase II and that we continue to see in the Phase III study so far based on the open-label 100-milligram arm reduction in fibroScan, MRE, fibrosis biomarkers, et cetera. I don't know, Stephen, if you want to add anything on that?

Yes. Just a couple of quick comments. When you look at Phase II trials, you really are trying to get at, is there a signal? And one of the things that Madrigal did was really dive into AI digital pathology early on and analyze those biopsies from the Phase II. And it's interesting when you look at an ordinal scale the way that semi-quantitatively that a pathologist looks at these biopsies, it's difficult to show changes in the margins of stages, where that is much better illustrated with fully quantitative assessment of collagen. And therefore, when we look at that data, we saw a significant improvement in F3 disease relative to placebo when treated with a generally lower dose of resmetirom. And so when you build your Phase III program and you're allowed -- you have the ability to really focus your power calculation on illustrating that benefit histopathologically, particularly on fibrosis. And I think that was very helpful in designing the MAESTRO-NASH trial. The other thing I would say is when you look at the MAESTRO-NAFLD-1 data, all the biomarkers are moving in the right direction. One of the biomarkers we historically look at for impact on histopathologic changes, ALT and the data that Rohit Loomba has published and all of us have kind of looked at in many of the trials with NASH, is a change in ALT of 17 units per liter. And here, you're seeing from data I presented today, at least a 20% -- a 20 unit per liter change from baseline. And so it's not one particular biomarker that's moving, it's collectively all of them moving in the same direction. And I think that gives us encouragement that we're on track to have a very good readout at the end of the day.

And our next question comes from Yasmeen Rahimi from Piper Sandler.

This is [ Jesse ] on for Yas. Congrats on the data and building off of Liisa's question. Can you please give us the cutoff for PRO-C3, ELF and MRE and the delta differences that were linked to fibrosis benefit? And I have a follow-up question to that after.

I'm not sure...

Could you repeat it and speak up a little bit, please?

Yes, sure. Were there cutoff for PRO-C3, ELF and MRE that were linked to fibrosis benefit? And what were the delta differences?

Yes. So we I'm not sure if you're asking generally whether there are cutoffs that are linked to fibrosis benefit on biopsy. And I would also invite Stephen to comment on this. But to my knowledge, what has been shown is that there are some predictors of progression with ELF -- with higher ELF. That in other words, to my knowledge, there's no direct correlation between a change in biomarkers and a change in fibrosis stage on liver biopsy, there's some early data with FibroScan reduction at the level that we're seeing with resmetirom, that it is correlating with a 1-stage reduction in fibrosis. But remember, there's very little serial data in studies, in which there is a fibrosis reduction. So Stephen, do you want to comment on any on that question?

Yes, I think that's -- Jesse, that's a good question. You can relay that to Yas, the results -- our comments from today. I think this is early days in some of these NITs. And 2 things stand out to me. Number one, there have been very few studies with paired histology that have gone so far as to look at PRO-C3, ELF and MR elastography change. In other words, do the liver biopsy baseline, do it at the end of the trial, but also get all 3 of those biomarkers along the way. Very few have done that. As a result, looking at the magnitude of effect change for PRO-C3, ELF, the components of ELF and MR elastography are not well-known. They're not well-defined. They're better known, at least from a ELF perspective and an MR elastography perspective for their prognostication ability. In other words, their ability to say, this -- if this is your score at baseline, this is the likelihood of progression of disease. Now that's the beauty of Madrigal studies. We have that data. MAESTRO-NASH will have that data relative to all 3 of these biomarkers paired with histology. Now what we do know, in small trials, particularly on trial done by Arun Sanyal and another one done by is that the percent change in PRO-C3 of 20% and the absolute change of 4-nanogram per ml in PRO-C3 have been associated with improvements in fibrosis. There are some early days, early data from some of the work done by Gilead linking about a 0.5 point change in ELF score with improvement. But the problem is, some of these trials early on were done with drugs that had minimal histopathologic impact. So it's hard to fully appreciate the magnitude of effect change, you're likely to see with a drug like resmetirom, a THR beta -- THR receptor beta agonist. So we're not sure necessarily that these biomarkers are agnostic to mechanism. They might actually be linked to mechanism a little bit like MRI-PDFF is.

Great. And then a follow-up question. For those 3 NITs, PRO-C3, ELF and MRE, were these trapped at week 16 through week -- at week 16 and week 52 for the open-label cohort, and how much more reduction was seen between those 2 time points?

So first of all, we did not show PRO-C3 data. At this point, we collected PRO-C3 data. We are using a commercially developed PRO-C3 assay by Roche, and that we're beginning to collect those data. So at this point, we haven't shown PRO-C3 data from the open-label arm or other points of the study. We showed some ELF data. The ELF data was collected in a time course throughout the 52-week and that was shown on one of the slides for the open-label arm and will also be available for the 52-week blinded arms. And for MRE, what we did show the data, where we had a significant reduction in MRE in the open-label arms. And those data will also be [Audio Gap]

Operator, I think we may have lost Dr. Taub.

Hello.

There you are.

We lost you for a little bit.

Yes, I'm not sure where you lost me, but I commented on the PRO-C3 assay. Did you hear that?

Yes.

MRE reduction.

I was just saying for MRE and for ELF, the MREs were done at week 16 and week 52 was statistically significant in the open-label arm of the MRE reduction, that was shown by Dr. Harrison today and also at the meeting. It was significant at both 16 and 52 weeks. The ELF data that we showed here is at the end of the study and also showed significance throughout the study. And then I also wanted to point out that the M30 or CK-18 data, which is a biomarker that's been associated with ballooning, was highly significantly reduced as well.

And our next question comes from Ed Arce from H.C. Wainwright.

Great. And congrats on this data. I wanted to focus on liver volume since this is clearly one of the new pieces of data from this analysis, and this is either for Dr. Taub or Dr. Harrison. Could you perhaps describe what we know to date in terms of how the reductions in liver volume are correlated or our view to be positive to liver fat reductions and in turn, either NASH and/or fibrosis? And is the goal with liver fat reduction a certain threshold? Or is it viewed, more importantly, to be normalized? And then finally, around liver fat -- excuse me, liver volume, could you also explain the importance of MRI-PDFF liver volume adjusted and explain how that is tied into the rest of the data?

Yes. So first of all, I say that these are new observations that we -- what we've shown so far is that there's a potent effect on liver volume by resmetirom that appears to be correlated with liver fat reduction as measured on MRI-PDFF, but is over and above what is predicted by -- simply by MRI-PDFF reduction. And we will be doing additional studies in these populations to understand what exactly the liver volume reduction is compared to potentially other mechanisms we've seen so far, that it appears to be more pronounced reduction in liver volume than by mere weight loss in placebo that this has been published or even by a GLP analog that was shown at AASLD. So Resmetirom was having more a pronounced effect on resmetirom -- on liver volume and a couple of other mechanisms that have been looked at, and we will be looking at that in more in detail. The point about reducing liver volume and then correcting for how much liver fat was reduced is not an easy conceptually, but remember that PDFF is a measure of liver fat concentration in the liver. It's not a direct measure of liver fat, but it's based on a concentration concept. So if your -- let's say, your liver fat at baseline is 10%, and that's based on a concentration in the liver. And now you reduce your liver volume by half without changing your liver fat, now your PDFF percent will be 20%. So you doubled -- you haven't changed the amount of fat, but you doubled what would be measured on the PDFF. So what we're saying is we use liver volume, then you need to correct for the PDFF reduction, which is actually higher, the liver fat reduction is actually higher with resmetirom by about 10%. And is shown by simply doing the relative reduction based on MRI-PDFF. And that's why we have a corrected liver volume of 61%. I don't know, Stephen, if you wanted to comment more on liver volume and your thoughts on that?

Yes. I mean -- I think that's an astute observation and question that you had. I mean this is certainly early days, but it's a novel finding that I think deserves significant further exploration and linking it to spleen volume, I think, is also going to be important. But all of these analyses, we plan on continuing to work on and we'll have further data in the future relative to this. But I think it's important to note, and I showed this in a different data set, that the change in liver fat content is -- while there is a correlation with weight loss, change in HbA1c, change in ALT, that there are other drivers of liver fat content reduction outside of just what we typically think about. So it's a novel area. It deserves further exploration, but it's exciting to see that we might be able to link this to different measures of response and potentially outcomes. We just have to wait and see.

And our next question comes from Ritu Baral from Cowen.

Two questions, a quick one. Dr. Harrison, you mentioned the sort of 2-step diagnostic process as you see real-world NASH treatment evolving in the future, this high negative predictive value PCP level test and the high positive predictive value referral center test. Is there any favorite that you have for that first step that high negative predictive value test that can easily be done at the PCP level with the standard diagnostic tools they have at a primary care office?

Yes, that's a great question. Again, in my post-grad course talk, I showed a picture of a biplane and then I showed a picture of a fifth-generation strike fighter. And we're at the Wright Brothers stage of these first NITs. I think those are going to evolve rapidly over time. But there's an algorithm we published with the AGA, American Gastroenterological Association, recently through a combined effort of different specialties, endocrine obesity, primary care and GI. And there, currently, while not originally designed for this use, it turns out FIB-4 at the primary care level is a test that potentially could be used to exclude patients that need further evaluation at least in the short term. Where you have Fibroscan capability, that's also a test that could be used at the primary care level. And where we use that would be a cap of 280 or less and a kPa under 6 or 7-kilo Pascal. And there are also proprietary biomarkers that you could order that could also be helpful at excluding disease. But I think at its basic pragmatic level, FIB- is free and it's relatively easy to calculate and then Fibroscan would be an alternative if you had access to that.

Got it. And then my follow-up is actually also on the liver volume point. Becky, the whole conversation around liver fat, adjusted liver volume begs the question, what else is driving decreases in liver volume? And for Becky and Dr. Harrison, if it's not just fat, is it -- is the delta additional fibrotic volume? Is it sort of inflammatory infiltrate? And for Dr. Harrison, are there any sort of quality of life sequelae of liver volume? I know that other prior companies in the NASH space have mentioned that abdominal pain may correlate with this and there were anecdotal reports of that in NASH patients. Any thoughts on that?

Yes. So I wanted to just say that the PDFF, as I mentioned, is an indirect measure of liver fat. It doesn't measure all forms of liver fat. So liver fat is still on the table. For example, we won't measure cholesterol, which resmetirom also reduces. There's also potential for, as you mentioned, inflammatory milieu, et cetera. There may be an increased pressure in the liver. There's some early data that suggests that PDFF reduction and spleen volume reduction, they go together, which will be very exciting, of course, if we can show that in the cirrhotic population. So hemodynamic types of things. And one other point I wanted to make was your QOL question. Because we have published that resmetirom in patients who did have reduced liver fat, does improve physical well-being, including bodily pain. And this publication was spearheaded by and came out this year. So that's one of the things that we also want to go back and look at and see whether the patients with -- there was also an association with the liver volume reduction. This was from our Phase II study. So good question on that. And then Stephen, I'm going to hand it over to you to discuss other potential reasons for -- hypothetical reasons for liver volume changes.

Yes. I mean absolutely. And just to point out that Glisson's capsule that surrounds the liver, is where the nerves are. So really the pain associated with the liver is when you stretch the capsule. In fact, you could apply that to anything in the gut. If you stretch it, twist it, pull it or -- that's where you get abdominal pain. So it makes sense that if you're reducing liver volume, you're shrinking the capsule, you should theoretically have less pain. Just again, very theoretical here. What's changing liver volume? I think it's early days. There's a lot of interesting concepts that could be looked at. One is blood flow and just hemodynamic circulation. And I think additional work that Rohit Loomba worked on looking at spleen volume, and showing that as you reduce spleen volume, you actually have a change in platelet count as well as change in liver chemistry test, et cetera. And I think that's not necessarily completely different than change in liver volume. As we begin to improve liver volume and liver fat content, there might be changes hemodynamically that could be contributing as well. It's early days. There's a lot more work to be done here. One of my areas of interest is going to be looking at these changes in organ volume relative to outcome measures, whether those be histopathologic or long-term outcome measures. I think both offer a huge opportunity to investigate. And we have that data within the MAESTRO-NAFLD-1 and NASH programs. So it's going to be exciting to mine that data over time.

And our next question comes from Jon Wolleben from JMP Securities.

One for me on the subgroup analysis you presented. Wondering if there's any common characteristics in the patients with low sex hormone binding globulin? And what percentage of those patients were of the 171 that completed the study? And if that was a smaller percentage than what you saw in Phase II at the lower resmetirom doses?

Yes. So it's not that they have low sex hormone binding globulin. It's the percent change in sex hormone binding globulin from baseline. So this is a -- sex hormone binding globulin is very specific biomarker for resmetirom and it increases -- it's a protein made in the liver. It's a neutral protein without a functional correlate, and it increases when -- with resmetirom. And it's an indirect measure of exposure of resmetirom in the liver. And so what we did was we arbitrarily had a cutoff where we took the 2/3 top sex hormone binding globulin responders and then looked at what their PDFF reduction was and what their liver volume reduction was. And you could see that clearly having that response was -- showed a more robust effect. So you could have a low response for a variety of reasons. But the -- and there are low responders that have very good results as well. But the obvious would be lower exposure to the drug, whether -- for whatever reason, that's really what that is. And there's no characteristic of the patient that would show that in any way.

And our next question comes from Eliana Merle from UBS.

Congrats on the update. Just for us kind of to follow up on the liver volume conversation. I mean, any plans to study in cirrhotic patients and just what this liver volume responder analysis could mean on a biopsy basis? Just sort of -- the significant sort of reductions in the liver volume? And then, I guess, just second kind of, coming out of AASLD, just what's your latest perspective and strategy is for serving potential combinations? And just, I know the NAFLD study was more of a safety database study, but just after, I guess, the top line and NAFLD will be, I guess, later this year or early next year, kind of your -- any potential plans to start further studies in the NAFLD population, particularly thinking kind of combinations?

So in the cirrhosis patients, we only had the 16-week liver volume because that's an open-label arm that started later than the non-cirrhotic open-label arm. So we'll be interested in completing the liver volume measurements along with all the other week 52 measurements in that group. And as Stephen mentioned, we're also interested in looking at spleen volumes in both non- and cirrhotic NASH. The spleens are also visible on the MRI-PDFF that we do. So yes, we're looking for ways of noninvasively monitoring potential responses in cirrhotic patients. And we think that, that could be very important. And then in terms of combinations, one of the things that this real-life MAESTRO-NAFLD study affords us, is the fact that we are looking to have lots of combinations in this study. These are patients that are on every imaginable therapy, most NASH patients and NAFLD patients are on 10 to 20 different drugs. And you can see that we're beginning to look at resmetirom in addition to certain diabetes medications. And we showed our self-group of GLP and SGLT2, which patients are on that, and it represents about 20% of diabetics that are on these meds or slightly higher number of diabetics that are on meds, and there -- these are meds that are good at controlling body weight even though they didn't lose weight, they were on stable therapies. And you see that they have a very good response when resmetirom is added to those therapies. So we are immediately thinking about what is the background therapy. We know many of these patients are on statins and several other therapeutic categories that might have some benefit in patients with NASH. And so our first thoughts on combination is background therapy that these patients are already on for their underlying metabolic diseases and showing that what the responses are when we add resmetirom. And then as far as new drugs and development, we are -- we believe that we have an opportunity for first-to-market here. And so we're mainly focused on development of resmetirom but we have a careful eye as well on other drugs in development, and we think that resmetirom would be a good partner for really all of the different drugs that are in development. I know I also have some thoughts on that just in general.

Becky, you went through it very well. The -- I think as we move forward, it's important to show that specific mechanisms have what their impact is on NASH resolution fibrosis improvement with an eye toward additional extrahepatic benefits because we want to treat whole patient, not just the liver patient. And I think the role for combinations is clear. What we really need to target first is finding that backbone therapy and then we can build off of that. But the MAESTRO-NAFLD-1 database, as you alluded to, gives us an opportunity to evaluate in a real-world setting what patients are actually taking in addition to resmetirom, and if there's any subset analyses that could show us an added benefit or synergistic benefit of those drugs, particularly some of the SGLT2s or even GLP-1. So more to come there.

And our next question comes from Thomas Smith from SVB Leerink.

Congrats on the data. Just a couple of questions on our end. Can you remind us how you're evaluating the liver biopsies of MAESTRO-NASH? Are you using 2 pathologists to 3? And how are any disagreements being adjudicated? And then can you talk about some of the steps you're taking to minimize the biopsy variability in the NASH study?

Okay. So this is the can of worms. I know that Stephen has a lot of thoughts on the biopsy. We have discussed with the FDA, our plans around the liver biopsy review. And we are -- which -- in other words, we have agreement on how we're evaluating the liver biopsies, basic agreement on that topic. And we have said that we have 2 central reviewers. That is our plan. Exactly how we're evaluating differences and how they read the biopsies is not -- we haven't discussed that publicly, but it is a very stringent process that we're undergoing. I think one of the things that was clear from some of the data that was at the meeting, is that to have multiple, multiple reviewers; more than 2 reviewers. First of all, it's incredibly complex when you're talking about thousands of slides. But the other thing is that it requires that these slides are digitized as opposed to being read directly as glass slides. And in general, pathologists still prefer reading glass slides, which I think -- but we're talking about a big change in terms of passing digitized images to multiple different pathologists, and I'm not convinced that you get a better evaluation from that process. I think Stephen was involved in a study that was shown at AASLD when there -- I think there was something like 9 pathologists, even. I don't know. I'll let you make your comments on the liver biopsy.

Yes. This is that time point and where we're really at, I think, a critical point of, of really trying to understand how to minimize variability, stabilize placebo response and really understand how our drug is impacting NASH. And that really has come to light from the ALPINE 2/3 data I presented showing that the fibrosis change really didn't correlate well with the NITs that we're all moving in the same direction and didn't associate with the prior study. So I think it's really, really important, I think the FDA, what I would say is there's lots of ongoing conversations with the regulatory authorities around how we can stabilize responses relative to the variability as well as placebo response. I personally have had conversations with them as recently as this past week. And I think there's really -- it's not just one particular thing that you have to do. I think, ultimately, where we're able to use consensus reads and where we're able to bring along AI digital pathology, is kind of a companion to the pathologist along with uniform staining procedures, I think we're going to mitigate noise, sampling variability and intra- and inter-observer variability. And what I can tell you is that the Madrigal leadership team is working diligently to bring together the best way to adjudicate these liver biopsies at the end of the trial based on all the collective data that's come before. And it's like anything else. I mean, in my presentation at AASLD, I highlighted the drugs that haven't moved forward in drug development. And what lessons can we learn from those potential failures that will help us as we continue to move forward. And I think all of these things, looking at consensus reads, potentially having AI digital pathology as well, I think will be very helpful and will help to minimize the variability. And remember, it's not just the variability of the pre and post-biopsy, but it's also the instability of the placebo response. I mean if you look at the difference between elafibranor and a obeticholic acid in their Phase III trials, the response on fibrosis was identical with the high dose of each drug, with a placebo response that was different and made the difference in whether one move forward or one didn't. So I think all of these things, what I can say is it's being carefully thought out. All of these different potential avenues of looking at the biopsies are being scrutinized exponentially. And I think we'll have a very good plan in place with the regulatory authorities on how to adjudicate these biopsies when we get to that point in 2022.

Yes. I'd just add that 2 pathologists pathologists are the same ones that read our Phase II studies and had very good agreement and showed the same findings. And one of our thoughts is that to -- first of all, I think the sampling issue is much of an issue as the reader agreement. I think the reader agreement is quite reasonable given what they're looking at is very good on the fibrosis component, for example. And I think when you have a robust action of the drug and a large enough sample size, which is such as what we see in Phase III, and in some of the larger Phase II studies, that the endpoints are quite achievable. I think what we saw in some of the smaller Phase IIs was more of an issue. So in order to get around some of the issues having a larger study with a robust mechanism of action is really critical. And we believe resmetirom is in that category.

Okay. Great. Yes. And then just a follow-up, I just want to clarify exactly what data we can expect to see from the NAFLD study this year versus early next year? Sounds like you're expecting the 16-week MRI-PDFF and the 24-week lipid data this year. But are there any other data sets we can expect in that top line this year before getting the 52-week data next year?

Yes. I mean -- so I can tell you the team is working really, really hard to get all the efficacy data in place and get the biomarker assessments from 52-week, et cetera. So this is -- what I can tell you is that we will show what the data we have that when it's gotten to the point where we can freeze the data and where the release is accurate. And you're correct, that the early time points completed a while ago, and so we're confident we have those data.

And we may have some of the later data.

We may have some of the 52-week but that's -- that would be very early to release less than -- about 6 weeks from the time getting those data. Just to say, the study will remain blinded. And so this would be group data that we would release. We can't -- we don't unblind the study until we lock the database.

Understood. So unblinded data in early 2022. Okay. Understood.

And our next question comes from Mayank Mamtani from B. Riley Securities.

Congrats on the progress. So maybe just first question for the company and maybe specifically to the safety profile that we are seeing here. On one particular area of interest is the diarrhea and the loose stools, where it seems like you're tracking better than the Phase II data. So could you just comment on the pragmatic aspects of the study that are just, in general, useful for clinicians as they look to use the drug in real world? And also, like just clarify, if there have been any grade 3 plus AEs nontreatment-related or treatment-related in the open label so far?

So commenting on the -- what is a mild loose stools that attenuated at the beginning of therapy. First to say that we've done -- we've dosed 100 patients -- hundreds of patients in Phase I for similar durations at twice the dose, up to 200 milligrams. And we are not seeing -- we did not see diarrhea. So this is a phenomenon that seems to be in a very central population. NASH patients have a lot of GI types of AEs. And it's a mild phenomenon, not a -- it's not intolerable to these patients. We had the dropout rate, as you saw until the label arm, was very low, 1.2%. So we believe that patients are well -- it's not a limitation of resmetirom as opposed to some other mechanism of action that are associated with severe nausea and vomiting, and require dose titration over time. This is not at all in that category, just to sort of ...

When it occurs, it's transient.

And it's transient. So as far as the safety profile of the drug, we're looking forward to the blinded arms for safety data, really important to have placebo control when safety data release, but we could say that in the open-label arm that the safety data look very good.

We have any grade 3 AEs?

So we'll be releasing 8 profiles across all the AEs -- all the AE categories at that point.

Great. And maybe just quickly for Dr. Harrison. Just a follow-up on -- so Dr. Harrison, as you know, there's a lot of work ongoing publishing the correlation of PDFF with outcomes and also MRD with outcomes. You and Dr. Loomba are pioneering that effort. I think you mentioned there is some publication also for liver volume with outcomes. Just quick status update on that? And then the second question I had was, are you -- could you comment on the overall profile that you're seeing for resmetirom across cirrhotics, non-cirrhotics? And putting in context, other developments in this space, forthcoming, for example, the Intercept's REVERSE data, how could the company see this broader label to be helpful for more advanced disease patients?

Yes. So that's a mouthful to go through. But I think to your first question about liver volume and outcomes. We don't have any data yet, Mayank, that's early days. We're in a very nascent time period relative to that. I did do a lit search this morning looking at liver volume and there is a paper in the American Journal of Radiology published this year looking at liver segmented volume ratio on CT scan, which shows a correlation of liver segmented volume ratios in NAFLD with advanced fibrosis. But it is not linked to an outcome measure yet. So further analysis to be done. There is additional work being generated on spleen volume and looking at response to therapy and particularly change in PDFF, and that's work that Rohit presented at this meeting. More specifically and maybe switching to your other question on cirrhosis. I think it's important to understand the significance of the MAESTRO-NAFLD-1 data. And I think kudos to Madrigal for looking at this phase -- a separate Phase III trial in a large number of patients using noninvasive technology, but also bringing in a broader range of patients. Some of these compensated cirrhotic patients exposing them to drug. I mean this is on top of Phase I data looking at hepatic impairment studies, where we're looking at PK data relative to severity of underlying fibrosis. This is actually giving patients drug for a prolonged period of time and monitoring them from a safety perspective and from an efficacy perspective relative to NITs. And that amount of data is going to be incredibly helpful when -- and if there is a filing that occurs, but also more importantly or maybe just as importantly, as we begin to look at alternatives, additional studies for resmetirom this patient population that we're studying already in open-label format. And looking at these NITs, we will be able to glean data relative to the -- to what's happening over time with drug in these populations that then may allow us to do additional studies in these populations. So I think it's -- it's early days, but that broad breadth of data that we're generating to it's going to allow us to mine that data and apply it towards additional patient populations and studies in the future.

And our next question comes from Matthew Luchini from BMO Capital.

Great. if I may, 2 for the company and then 1 for the physician. First, to the company. I guess just a clarification on the NAFLD study. Can you clarify when we should expect any color on safety? Would that be part of the 2022 release? Or should we expect something this year? And then understanding that the question, the sort of reader variability strategy hasn't been disclosed as of yet, is that something that would be disclosed prior to data release or not until that point? And that's obviously in relation to the NASH study. And then after that, I do have one for the physician.

Yes. As we've said this entire year, we plan to release the safety data at the beginning of 2022 after the database lock occurs. We are well aware that we still have patients in the study, they finish at the end of this month. And so therefore, we won't expect the final safety data to have been collected until the end of this month, and then we'll read out at the beginning of 2022. So this is not something we're only saying today. We've been saying this the whole year. And there are no shifts in our timing. As far as liver biopsy strategy that we're using, we believe that Dr. Harrison alluded to, that this has been a rapidly evolving concept. We believe that we have a very strong approach and we have a sound approach to evaluating the liver biopsies, don't expect changes in the strategy to occur. However, because this is a statistical analysis plan that is negotiated with the agency. It's not the sort of thing that we would be expecting to announce in detail publicly. It's -- I don't know...

We'll provide more granularity on what's happening as we move along. But we may not get into every detail of the reads until we have the data.

Okay. Understood. And then just quickly for the physician, coming back to the PCP-oriented diagnostic algorithm that you mentioned earlier. Would just be interested in your perspective on how we should be thinking about basically the dissemination of that algorithm into the PCP community by the time this Madrigal might be potentially launching? In other words, would we think that this would be something that would be kind of well understood and relatively broadly used or sort of still in the early rolling out phases by the time the product would actually launch?

Yes. I mean I think we focused our efforts early on in trying to find drugs that were impactful in this liver disease. And now as we're on the precipice of really kind of seeing what impact these drugs are having on NASH. And as Paul mentioned, getting to a potential for a new drug application, it's important we begin to also focus on disease awareness. And it's not just disease awareness at the physician level or at a specific specialty of physician level, it's really getting after those grassroot physicians that see the predominance of NASH and NAFLD. And it's also getting at the patients and building disease awareness at the patient level so that they ask the primary care provider about their potential risk for underlying NAFLD or NASH. So it's really a combined concerted effort in trying to drive disease awareness. And in so doing, we need to be very prescriptive and how we do that. We don't need to be confusing providers with 52 different algorithms and a 1,000 different ways of assessing disease severity. We need to begin to coalesce our thought around what biomarkers we can hang our hat on despite the fact that they're not honed and refined yet, what do we have in our toolbox that we can apply that is broadly applicable to that level of care. And I think right now, we're at the point where we want to have a biomarker that could be done relatively cheaply, if not free, and that it could be -- is composed of routine chemistry test that the primary care provider gets in. FIB4 gives us that. This is not a fantastic test for NASH. It is developed in co-infection, HIV and hep C to identify at-risk fibrotic patients. In the field of NASH, it was developed to identify NASH with F3 and F4 fibrosis. However, what we do know in additional analyses is that it has a very high negative predictive value for excluding bad disease, and it's composed of ALT, AST, platelet count and the patient's age. So these are variables that are very easy to obtain at the primary care level. So I think starting with that, that's at least what we've said in the AGA publication on our algorithm for management would be the way to go. If you do have, as I alluded to earlier, fibroscan, that also has a negative predictive value for excluding at-risk NASH of over 90%. And I think that's kind of the basic level. There are tests in development now. There's NIS4. There's FAST. We're putting together MAST. There's method to identify prospectively with high positive predictive values, those patients that need to be -- that are at higher risk for progression of disease and would then warrant therapy. And I think that's also being -- That's also an area that needs to be focused on and developed a bit more. But I think that -- hopefully, that answers your question in a little bit more detail?

Yes.

And next question comes from Liisa Bayko from Evercore ISI.

Sorry, just one follow-up. I wanted to clarify, and I know you've been saying and I am aware of that, but you're going to be giving safety data more in the early part of 2022 time frame for the MAESTRO-NASH study. Can you maybe just describe like what kind of details we're going to get at that point? I'm assuming it would be broken out by ], and it would be someone unblinded. I heard something about it being not unblinded. So anyway, I just want to clarify exactly what we'd be getting that?

We're going to release top line data. We expect to have very -- a lot of the details provided at a major meeting presentation. We'll be targeting probably a late-breaker at EASL, which is in April of this year. But to your question, we will provide the basic AE breakdowns as far as safety goes and some of the key safety-related endpoints as top line.

And our next question comes from Steven Seedhouse from Raymond James.

Dr. Harrison, just wanted to ask about the background thyroxine treatment. I think 44%, if I caught that in the baseline table and MAESTRO-NAFLD-1 open-label were on thyroxine. I want to ask first to the enrollment algorithm either directly or indirectly in patients on that? And there any patients that weren't on it at baseline started after the study treatment began? And then also, just with respect to the upcoming placebo-controlled readout, what's your take on how this would affect either thyroxine toxicity in the placebo arm? Just trying to understand how we should interpret the placebo-controlled data that as background?

Yes. So the question relating to thyroxine treatment. First, let me tell you that resmetirom has no impact on the thyroid access. We published this in our NASH Phase II study. It had a minimal effect on free T4, about 11%. This is a pro forma, which is then converted into the active hormone in the liver, the active hormone being free T3 in both of our MAESTRO studies allow patients to be on thyroxin. There's a significant fraction of NASH patients who are clinically hypothyroid at the level of thyroid gland. So it's typical to see in -- on the order of 13%, 15%, 20% of people on thyroxine. In the open-label treatment arm, we specifically enrolled a number of people on thyroxine so that we could demonstrate the fact that resmetirom is safely co-dosed with thyroxine. So there's no necessity for resmetirom to be dosed with thyroxine. I think in blinded arms and MAESTRO-NASH study, it is around 13%, 15% of patients on thyroxine. But we did have more patients on thyroxine in the open-label noncirrhotic arm and we will be showing some subgroups to show that there's no difference in terms of response to resmetirom in patients on or off thyroxine.

Okay. If I could just ask Dr. Harrison as well. I think in some other studies, you've run in NASH, you've either treated or prophylaxis patients with antidiarrheals or antiemetics. I just wanted to see if that's happening in MAESTRO-NAFLD or MAESTRO-NASH? That's the second question.

No, it's not.

No. We -- again, this drug has been well tolerated in clinical trials, both the MAESTRO-NAFLD-1, MAESTRO-NASH in the Phase II trial. And that's been my experience in the patients I've enrolled as well.

Okay. Well, we've run a good deal later than we had originally scheduled. So I think we should close out the Q&A at this point. I'd like to thank everyone for joining us today and hanging around for extra innings. I especially like to thank Dr. Stephen Harrison for reviewing the data. We recently presented at the Liver Meeting. The data are encouraging as we look ahead to the planned releases of top line placebo-controlled double-blind data in the coming months from both of the 2 studies, and we look forward to speaking to you again. And with that, operator, let's end the call.

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