Alnylam Pharmaceuticals, Inc. (ALNY) Earnings Call Transcript & Summary

Welcome, everyone. Glad to have you here at HFSA. This is Alnylam's first in-person investor event post pandemic. So we're really extra pleased to have you here as well. I'm Christine Regan Lindenboom, I am the Senior Vice President of Investor Relations and Corporate Communications at Alnylam. And we've got a great panel with us today. So we'll have Pushkal Garg, our Chief Medical Officer, give some opening remarks. Matt Mauer is joining us from Columbia University, and he will review the data that was presented earlier today. And then we'll go through some summary next steps. We'll then have Tolga Tanguler, our Chief Commercial Officer, join us for Q&A as well as John Vest, our Vice President of Clinical Research as well. Before we start, I wanted to share some forward-looking statements, please review those at your leisure. And then we will take Q&A both from the room as well as from the webcast, if you're on the webcast and you have a question, feel free to type it in ask a question box, and we'll try to get to as many as we can. And with that, I'm going to turn it over to Pushkal to lead us off.

Thanks, Christine, and thank you all for choosing to spend your Friday evening with us. We really appreciate that. It's really exciting to be here at HFSA sharing additional data from the APOLLO-B clinical trial of patisiran in ATTR cardiomyopathy. I'm going to turn it over to Dr. Mauer in just a few moments to actually walk through the data that were presented today. But maybe at a high level, I just want to give the perspective that we couldn't be more excited by the results. Building on the top line data that were presented at ISA about 3.5 weeks ago that showed statistically significant and clinically relevant benefits of patisiran on 6-minute walk test and KCCQ. The results today look at those data in -- across a number of prespecified subgroups and show a general consistency of effect across all of the prespecified subgroups. In addition, important data from exploratory analyses indicate benefits or favorable effects on measures of cardiac stress, cardiac injury, cardiac structure and cardiac function coming from biomarkers and imaging modalities. So all very exciting and building upon corroborating the top line results that were presented a few weeks ago. And that, along with an encouraging safety profile together really fuels our belief that patisiran and RNA therapeutic agents in general, that target TTR can be effective therapeutics for patients with ATTR cardiomyopathy. So -- that's important as at Alnylam, we are looking to build an industry-leading TTR franchise. If APOLLO-B, we will be submitting it for an sNDA at the end of the year. And if that's successful and meets regulatory approval, it may be on the market at the end of 2023. It builds upon our already 2 commercialized products, ONPATTRO and AMVUTTRA for the polyneuropathy of hereditary ATTR amyloidosis. And beyond that, we're looking forward to the HELIOS-B trial vutrisiran in ATTR cardiomyopathy. We're kicking off a program with ALN-TTRSC04, which has the potential to be a once annual therapy for patients with ATTR amyloidosis as well as exploring the potential of TTR silencing to be a therapeutic for Stargardt disease, an inherited retinal disease. So together, we think we have the potential to build a franchise that maybe offer a tremendous benefit to patients and also as well as be a big source of growth and value creation for the company. And all of that feeds into our aspirations for our P5x25 strategy, which is a vision to actually have a top-tier biotech that is delivering medicines for rare and prevalent diseases to patients all around the world, advancing the clinical development of innovative therapeutics that can transform the lives of patients, built upon our organic product engine and as well as deliver strong financial performance. So with that, I'm going to turn it over to Dr. Mauer, who's going to walk you through some of the data that were shared today. Matt?

It's really a pleasure to be with everybody. I must say these data look better and better all the time. I have to say, the more endpoints we look at, the better they look. So just to review, I think you're all aware, transthyretin cardiac amyloidosis is a pretty terrible disease. It's rapidly progressive and it results from a misfolded monomers or ligamers of the transthyretin protein that can deposit in many organs, but the heart is one of the main ones. And patients can either have a hereditary disease inherited in an autosomal dominant fashion or wild-type disease. It results typically in a restrictive cardiomyopathy with a heart failure as a manifestation. Obviously, in heart failure patients have declines in functional capacity and quality of life. They have increased hospitalizations and unfortunately, pretty short in survival. And patisiran, which we've been leveraging for patients with transthyretin amyloid polyneuropathy for several years since its approval in 2018, is IV-administered RNAi therapeutic. And the prior data that we analyzed with Scott Solomon and others, and we published several years ago showed that in patients with polyneuropathy, who have concomitant cardiomyopathy, there were demonstrable effects of patisiran on important cardiac endpoints, biomarkers and echocardiography. I would say, parenthetically, if I had one thing to do over in the field, I would probably have a few I might have not led us to dichotomize into neuropathy and cardiomyopathy because many patients have both. So APOLLO-B, as you, I think, know, is a Phase III randomized, double-blind international placebo-controlled clinical trial in patients now with transthyretin amyloid cardiomyopathy. And there were 360 patients enrolled in this trial. They either had wild-type or variant disease. They had a symptomatic heart failure. There were key entry inclusions related to biomarkers I mentioned previously, their NT-proBNP had to be above 300. If they're in sinus rhythm or greater than 600 , there was an upper limit of 8,500 for certain patients and that have a minimum walking distance. And up to 30% of the patients were allowed to be on background tafamidis therapy for at least 6 months and in the opinion that the investigator were progressing. Subjects are randomized 1:1 to either patisiran, which was given IV every 3 weeks, along with concomitant medications to prevent infusion reactions or placebo. The placebo group also received those concomitant medicines. And randomization, as you see at the bottom here, was stratified based on several variables. Importantly, whether they're taking tafamidis, yes or no. Do they have a hereditary or wild-type disease because progression is very different. And do they have a New York Heart Association Classification I or II disease and were they less than 75 versus all others. The primary prespecified endpoint was the change in the 6-minute hallwalk distance from baseline to 12 months. Key secondary endpoints are shown here. The KCCQ is universally used in cardiomyopathy trials, and we use the overall summary score, and it's changed from baseline to 12 months as a measure of health status or some people say quality of life, and we looked as well at death and hospitalization outcomes over the 12-month study period. And at the meeting here, we presented -- my colleagues presented data on the cardiac biomarkers and very intriguing and positive reflects on imaging. And the patients are now in an open-label extension, which is fantastic. So these are the high level, if you will, Table 1 data that demonstrate randomization works. The groups are well balanced across all the parameters. As you can see, the average age of these patients is 76. So male-dominated disease, which we've always struggled with a little bit, but it's still borne out even in our clinics. 80% had wild type. You can see down that their average NYHA Class II baseline tafamidis use was 25% in each group and the Gilmore staging system that we talked about earlier, using NT-proBNP above 3,000 or an eGFR less than 45 is used to rate patients into Stage I, II or III disease. And you can see most patients had Stage 1 or 2 disease. 6-minute hallwalk distance, about 360 meters compatible with Class II symptoms. And so was the KCCQ overall score of about 70 with median NT-proBNP levels across the both groups about 1,900, I say, picograms per ml, but here is nanograms per liter same units, but depending upon whether you're in the states or across the pond, as they say. So patisiran is really very effective at its therapeutic goal of lowering TTR. And you can see here, rapid reductions within 3 weeks. At month 12, there was an 86.8% reduction in serum TTR in the full cohort. And you can see that there were comparable serum reductions in TTR irrespective of whether you received tafamidis at baseline or not, and we're taking it through the duration of the trial. So that's important. This is the data from the primary endpoint. So it demonstrated a significant clinical benefit. As you can see, the kind of group in red, if you will, is the placebo group, and they have a steady and progressive decline measured first at 6 then 9 and 12 months with a change in their 6-minute hallwalk distance on average of 21 meters declining. And the group on patisiran is quite stable. And the absolute here to say the median difference was 14, 15 meters and statistically significant. And we showed, I think, at the ISA in Heidelberg that the prespecified sensitivity analysis using the MMRN confirm these findings with a similar statistics, a slightly greater effect size. These are the data for the secondary prespecified endpoint. That was the quality of life of patients with their KCCQ. And again, I think you can see remarkably stable self-reported quality of life based on the overall summary score over the 12 months that unfortunately doesn't occur in patients who were taking placebo. And these differences were nearly 4 points with, again, a statistically significant p-value. These are the new data that we've presented, I think, here, and this is Forest plot showing the effect size across many clinically relevant cohorts. And as you can see, overall, the point estimates and the confidence intervals and some are wider, but the point estimates always favors essentially patisiran, both for the 6-minute hallwalk distance and for the KCCQ overall summary score. So very, very, very consistent benefits across multiple different subgroups. This is the time to first event and these events include kind of harder things, if you will, all-cause hospitalization, urgent heart failure visits or deaths. And you can see for the full analysis set, a hazard ratio that's not statistically significant across as one but favors patisiran as opposed to placebo and subgroups divided by whether they were taking baseline tafamidis or not on tafamidis, someone asked in the previous, without these differences, and I think these are very small subgroups, so I would interpret them with a lot of caution. I think there are 91 people in the entire cohort that are taking tafamidis that can divide it up into little groups. So all-cause mortality over the 12-month double-blind period is shown here. And again, in this trial, as in others, when patients progress and are sick enough to, god forbid, need and are eligible for a heart transplant or rarely, if they are sick enough and they're eligible for LVAD, we equate that to equivalent to death. That's been the standard in the field. And you can see here that in patients on placebo, there were, unfortunately, 10 deaths and only 4 in the patients on patisiran. 2 of the subjects got a heart transplant on placebo and none on patisiran and the CV deaths were 5 and 2, you can see, again in favor of patisiran. The hazard ratio estimate, again, crosses 1, but it's certainly way in favor of the drug with a point estimate here at 0.355. And the data here doesn't differ whether someone is taking baseline tafamidis or not. Numerically, in all circumstances, patisiran has a more favorable hazard ratio than placebo in the subgroups, as you can see. These are also, I think, a new and somewhat complementary data. This is the just the geometric fold change in NT-proBNP, a marker of cardiac stress. This rises quite a bit in patients. I recall back to our natural history study years ago, People's NT-proBNPs, you know, went up 1,800 picograms per ML in the TRACK study years ago that defined the natural history in 6 months. So it's a pretty bad disease. We've obviously identified people now who have an earlier phenotype, but the increase you can see in NT-proBNP over a year and the placebo arm was on average from baseline 518 picograms/mL and nanograms/L, pretty significant increases. And you can see that patients on patisiran really had a very minor increases, probably not at all clinically significant. And here, the statistics aren't even like close to being borderline there like highly statistically significant. This is a similar data for another cardiac, if you will, stress marker. These are patients who have often what we call colloquially troponemia and they come into through an emergency room. They have a troponin elevation. They often get inappropriate catheterizations because people think in the absence of a coronary syndrome, they're having coronary artery disease, but this is a marker of the disease. And here, we're using high-sensitive troponins. And you can see, again, the mean full change, much lower in those on patisiran than those who were taking placebo. And you can see the baseline data was well matched for the high-sensitive troponin I and the changes there. These are also new data that are presented at the meeting just recently, and I think are beyond encouraging, really amazing at such a short period of time that we can see these, I'd say, personally highly significant and dramatic effects on cardiac structure and function. So you can see global longitudinal strain, which is a measure of myocardial shortening is getting significantly worse in those on placebo, and that's attenuated by treatment with patisiran. LV mass, which is composed ideally, the heart is composed only of muscle and intracellular tissue, but in this case, a lot of amyloid is very stable or slightly better. Declines in LV mass in those on patisiran, while those on placebo progress. Over time, the ventricle with this disease usually gets smaller, less capacious if you will. The end diastolic volume usually goes down, and you can see that happening in the placebo group, which is not happening with the blue, the patisiran. And as a result, if the ventricle is a little bigger and it can generate a little better stroke volume, then cardiac output is maintained, which is one of the central physiological responsibilities of the heart. I'm not touching a thing. I think it's just -- and the wall thickness also, which has been kind of an old fashioned, and I would say, not very robust measure from years ago, it basically also shows trends in the right direction as does what's called the relative wall thickness. Okay. Looks like we need some duct tape. And these are also new, and I would say, pretty provocative data. So as you all know, the diagnosis of transthyretin cardiac amyloidosis has also undergone a revolution. We used to have to stick people under a cath lab table and do an endomyocardial biopsy, and so we use technetium-labeled agents, different ones in the U.S. versus Europe, PYP in the U.S. HMDP also in the U.S., I recently learned, it is approved, by the way, both are available. I used to tell people in lectures, there's only one in the U.S., but I turned out to be wrong. And that because we just had a shortage of PYP, so we all found out that HMDP was also approved, and we started using, it works great and also DPD scintigraphy. So these are used to diagnose the condition. There's emerging data about quantification using the CT scans. In this particular subgroup, there were 37 patients who had serial scans with patisiran and 28 in placebo. So it's not the entire cohort. But as you can see in the chart, if you will, on the upper right, almost everyone is a grade 2 or 3 that makes the diagnosis, almost everyone's grade 3 to start, and there are patients who go from grade 2 to grade 1 or grade 0. And that's not something we expected to see, to be frank, in the world of amyloid, and I've never seen that with any stabilizer therapy. So with all the caveats, we've never known what Technetium is binding to, to begin with. This certainly suggests that something important and demonstrable is happening in the heart in a very good way. And there was essentially no one who got placebo, who got any better or had to change. You can see all the red in fact patients on placebo, probably, if anything, one of them got worse. So this is, I'd say, one of the more provocative things to come from this study, especially as someone who years ago said scintigraphy is good for diagnosis, but it's never good to follow patients. So I've said a few things that are probably wrong, but at least I admit them. So these are the safety data. I think mistakenly, people were worried about this drug in the past, but you'll see there's nothing to worry about. Majority of AEs were mild to moderate in severity. AEs that occurred more often in the patisiran group than the placebo, what everyone expected, right? An infusion reaction, or arthralgias or muscle spasms. Okay. But numerically, you'll see the AEs, including the important ones we mentioned at the bottom, deaths and were numerically in favor. And they changed a little bit because deaths related to COVID, which I would note, was only one person in the whole trial, which is kind of unbelievable in the midst of COVID, that this entire trial was conducted by Alnylam and the sponsors in a fashion, I would just say that is probably better than any clinical trial I've ever seen with regard to fidelity and to the proposed intervention is just a mind-boggling phenomenon. I mean, never mind that it's wildly successful and that every endpoint is that it was conducted in a way that allowed patients to get access to the therapy and get it at times at their home. So it was convenient and not be at risk for developing COVID is really a testament to not only the sponsor and everyone they work with, but also I'd say the patients. This is the cardiac safety data that just uses standard [ metric ] queries that clinicians provide. And as you can see, everything is less on patisiran than placebo; overall cardiac disorders, cardiac failure, arrhythmias. I would note no one had Torsades. That's just a reflection of the nomenclature that is used in this. But certainly no safety signal and if anything, a trend toward benefit. Maybe I just want to hang out there at that little spot. Okay. I'll keep going. So I'd say, in summary, this is a really short-term trial, right? I mean it's 12 months. I can recall times where KOL sat in a room, and obviously, sponsors always want to try to develop their drug as quickly as possible. And we all said neuropathic effects at 18 months, you'll be lucky and cardiac, you need to take a long time. So I think it's just amazing that these changes are seen in this relatively short time period. And it definitely supports the hypothesis that are RNAi therapeutics to reduce TTR. They nicely say potential treatment because that's probably the right thing to say, but I think they're going to be a treatment for patients, in my opinion. And that every primary and first secondary endpoint, there were consistent benefits as we show now across all prespecified subgroups a time to first event, including important ones of all-cause hospitalization, urgent heart failure visits and the god-forbid death and even all-cause mortality directionally, really strongly favored patisiran, and that was, again, only over a short period of time. And the exploratory endpoints that are presented here, as I said, are just more icing on the cake. -- cardiac stress, measured by NT-proBNP and troponin are all beneficial with patisiran compared to placebo. Most echocardiographic parameters move in the right direction, those that don't are probably just derived and have higher variants. And amazingly, scintigraphy seems to -- at least preliminary data suggests that things might be getting even a little better, certainly are stable. It has a very acceptable safety profile. There's been no emergent cardiac safety concerns. And thankfully, I've had the privilege of caring for my cohort, and they're still getting access to the drug and now they're all open label in an extension trial. So congrats to the patients, most importantly and to Alnylam.

Thanks, Dr. Mauer for that fantastic summary and presentation. We'll take some questions in just a moment. But maybe I want to do -- sorry, I was just -- very briefly just summarize next steps. So we're obviously very encouraged by these data. We are intending to file an sNDA with the U.S. FDA by the end of this year, hopefully, for approval. We are also looking into where other geographies and regions where there may be regulatory submissions subsequently. If this is approved by the FDA, we would anticipate being able to commercialize patisiran, ONPATTRO for ATTR cardiomyopathy towards the end of 2023. And then as we've talked about at the beginning of this, we're also eagerly anticipating the vutrisiran study in the same patient population, which is expected to read out in early 2024. And certainly, the results that we've seen here across the whole range of endpoints that Dr. Mauer just highlighted, really give us further encouragement in the probability of success of that study, which is a larger and longer study. So with that, I do want to just close by building on what Dr. Mauer said, which is that this study couldn't have been successful without the 360 people and their families who volunteered, it's a sizable commitment and a matter of trust that those patients volunteered for the study. We're very appreciative to them. We're appreciative to all the investigators and site staff, study coordinators and colleagues at vendors, CROs and Alnylam colleagues who made the study possible. So thanks to all of them. And we will open it up for Q&A. So I'm going to invite up Dr. Mauer. I'm going to invite up Tolga Tanguler, who is our Chief Commercial Officer; and John Vest, who's the Senior Vice President and leads our clinical TTR franchise. And am I supposed to sit up here, too? All right. Okay. So maybe we'll start. We've got a number of people in the room, and we also have people on the webcast. So we'll start with some questions in the room -- and if everyone could just introduce themselves, that would be great.

Ritu Baral from Cowen. I want to zero in on that figure 7. Dr. Mauer from your poster, the one with the subgroups and time to first event. Is there any biochemical rationale as to why one of those subgroups look different than the other, any biochemical rationale for synergy? And if you looked at those 2, I know there's small numbers, but it seems like the separation in the background half occurred in the second 6 months versus the separation in the -- without tafamidis group against patisiran was in like the first 6 months. So the delta was sort of in the first half versus the second half. And I believe you guys in your last call said that, that was sort of a cluster of hospitalizations that occurred in the tafamidis group early on. Is there any sort of rationale for that, biochemical rationale, because that seems to be the only figure that didn't align with most of the other figures presented across both posters today.

So maybe -- thanks, Ritu. So I think just for the folks in the audience, I think the question is really are around the time to first event on the composite endpoint and really just trying to understand if there's differences potentially biologic reasons that might explain why there might be different trajectories. But maybe what we could start is Dr. Mauer if you just want to start with just your overall gestalt as you look at that and if there is any biologic reason why you would think there would be -- how do you interpret those time to event curves in the subgroup, you started to talk about it in your presentation.

Yes. I mean I think the data is very robust, as I said, across a lot of prespecified endpoints, everything is in favor of drug. I think you can get into the weeds and I think we're really getting into weeds that are uninterpretable. And I don't know of any biological reason why one group that you're mentioning would do worse. I mean broadly in the field, one of the questions out there is will combination therapy be something that will be useful in some regard. There are debates about that, to be frank, if you knock out 90% of your TTR, I'm not sure what the role per se of a stabilizer would be in those particular circumstances. But my general belief is, I think you're asking too much of these data. We're talking about ends that are really, really, really tiny, not only in the number of people we're looking at, but also in the events. And so we all know this, things can flip one way or the other. So...

Okay. So my next question is going to annoy you. It's about even smaller-end subgroup. This was the New York Heart Association Class III on the whisker plot. That was the one that sort of deviated versus the others? Or do you think that Class III patients are just sort of beyond saving? Or how should we think of them as far as treatment benefit?

Yes. I mean I think your question kind of goes a little bit to the old ATTR-ACT data in which we showed across multiple prespecified endpoints, everything was consistent, but there was one that was not consistent. It was highly statistically different, right? It wasn't like just a little bit. It was that the patients who lived longer with NYHA Class III disease spent more time in the hospital. And that's predictable. This is a disease which is very hard to clinically manage. And if you pass the point of no return, you then god forbid have some dietary indiscretion, you develop an atrial arrhythmia, it doesn't take much to tip those people over the edge. I would just parenthetically say probably back in the day when we were doing ATTR-ACT, some of the patients who investigators claimed were Class III probably may have even had Class IIIb or IV disease because we're not unbiased as investigators. We really, really like our patients, and we do everything we can to try to get them into the trial, so that is the bias. I don't think that subgroups that you're talking about, again, I think we've all come to the conclusion that it's quite possible that silencer therapy will be favorable for people who have more advanced disease because the sense is, as you see from these data, it looks like it's stabilizing patients as opposed to allowing them to progress. And I don't want to read too much into the echocardiographic or PYP data or scintigraphy data, but I don't think stabilizers reverse this disease at all as their name implies, whereas there's the potential for silencers to really stop the disease in its tracks or potentially even in the long term, as Dr. Hanna was saying at the meeting today, maybe even reverse things. So...

Maybe just on this point, I think that subgroup that you're talking about is, I think, if I recall 28 patients, I believe, right? So it's 14 and 14. Maybe, John, do you want to just speak about some of the other indicators in terms of severity of disease like baseline 6-minute walk test and BNP and what that might tell us also.

Yes. Thanks. So as Pushkal is pointing out, that NYHA Class III group that you're talking about is 28 patients. But if we look at the other metrics in that forest plot that would be markers of advanced disease. So NT-proBNP is greater than 3,000 and baseline 6-minute walk tests that are less than 150 meters. Those are -- they're still small groups, but they're bigger than 28 patients, okay? So they're more adequately sized. And there, we see evidence of -- a clear evidence of a treatment effect that's consistent with what we're seeing in the overall population. So we look at that at least that gives us confidence that this is a fairly consistent effect across this entire population, including the full range of disease severity.

One last quick one, for Pushkal. What are your plans currently for releasing the 18-month open-label data?

Yes. So right now, these patients just crossed the 12-month time frame in the middle of this summer. So we're continuing to follow them. As Dr. Mauer said, we've got them in an open-label extension. All patients are on drug, and we're going to continue to collect the data. And as is our practice, we will release those data at the appropriate time, but we haven't made a specific determination when that will be. We'll talk with the investigators and do that in due course.

I'll speak out and say I'm following patients who've been on drug for more than 2 years now, and they're all doing perfectly fine, right? That's my own clinical -- that's not all 360 who were randomized and left over. But long term, the data, the patients are doing really, really well.

That you would use both silencers and stabilizers together in a patient population that you would treat. If so, what subset of the patient population do you think would be ideal for that? And then a couple of questions for the Alnylam team. Just on HELIOS-B, have you made a decision as to whether or not you will take an interim read on that because you did seem to reiterate on the slide just now that the data sets do, in the early part of 2024. And then last question related topic. You did make changes from APOLLO-B to HELIOS-B in terms of letting 50% of patients beyond background tap versus the 30. And you also took away the language that patients be required to advance while on tafamidis . Can you explain what led to that decision to make that criteria different in the 2 studies?

All right. So a couple of questions. I guess the first, Matt, if you could just speak about based on these data, that gives you any perspective in terms of whether you think combination use would be appropriate or not appropriate or how you think about that.

I don't think these data provide unfortunately, to us, clinicians, any insight about that, and that's going to be a holy grail. Unless you think that doctors make decisions about health care; no offense, you're going to have a hard time convincing the insurance company without hardcore data to give someone $750,000 potentially of therapy every year. So I follow currently more than 700 people on tafamidis and lots of people on patisiran, nowhere near the same numbers clinically, but I can tell you that I have less than 5 people who are on expensive combination therapy. That doesn't mean we don't often try to leverage diflunisal in patients who are currently taking patisiran much cheaper, but I don't think these data. And that's one of the struggles that I'd just say as a doc, we're all going to be faced with in the amyloid world is how to decide which therapies people should get and when they should get them and so forth, those are -- we anxiously await more information, but we're not running to prescribe combination therapy.

Yes. Tolga, do you want to speak from how you've been thinking about the potential places where something like patisiran may be used, obviously, subject to regulatory approval and what the label looks like.

Yes. I mean, as Dr. Mauer indicated, the label really defines how we're going to be able to communicate the benefits of this product once it's approved. Look, what we know based on the research that we've done and the advisory boards that we conducted is that there is a need, right? We also -- that's confirmed by the TRACK study. We know that there are patients who are progressing. -- and they're inadequately responding a group of patients that are inadequately responding. So that's one group that we believe is actually going to be eligible or we would actually position based on the label that we would get. The second one is, again this is based on what Pfizer communicates particularly in the United States, is those patients that get the prescription, but they don't end up actually getting the product due to a number of financial reasons. Part of that, we believe, is driven by Part B where the co-pay burden is directly on the patient. And in our case, as you know, we are physician-administered, therefore, it's Part B and the burden, especially those patients that are on Medicare, which mostly are, the supplemental insurance actually removes that burden for those patients. So that would be a benefit as well. And more importantly, obviously, we see this not just a sprint, but a marathon. And the fact that we have the HELIOS-B data coming out relatively soon. That will -- and with the outcomes data, that will obviously have a, I believe, a different position. But we're excited. I mean, look, we don't have a lot of products in this highly devastating disease, as Dr. Mauer indicated that sort of you have great clinical benefit and you have good quality outcomes and the cardiac outcomes, all moving in the right direction, a lot of good signals. So these patients do need alternative treatments. And because of that, we're really excited about that.

Thanks, Tolga. Maybe I'll handle your IA question Tazeen and then John, maybe in a second, you can address the third question that Tazeen had about the population enrolled in HELIOS-A. Look, with regard to the interim analysis, as we've said, HELIOS-B does allow us to do an interim analysis. And that is something that is -- will be deciding in the near future for the study. And we'll make that announcement when we're ready to do it. And obviously, we'll be discussing with that with regulators as well. As we've also said, the time frame because of how quickly HELIOS-B enrolled, the time gap between the 2 studies has become narrower, which is a good thing. And that would be a consideration in terms of not doing the interim analysis. But certainly, we want to -- these data are very encouraging. We want to make sure that the -- we can bring forward an appropriate medicine as quickly as possible. And so that's -- those are the considerations, and we'll get back and we'll be clear about that at the appropriate time. John, do you want to speak a little bit about Tazeen was asking about the entry criteria for HELIOS-B and the number of patients who might enter on baseline tafamidis and whether there's any implications to that in terms of the powering of the study and the success of the study.

Yes. Look, both of these studies when we designed them, what we're trying to do is to bring in a population of patients that's going to be reflective of the global population with this disease, and that's what we did. So at a high level, I just want to say that these 2 trials really are enrolling a very, very similar patient population. You've pointed out a couple of subtle differences in the inclusion/exclusion criteria, which are both accurate. But that really is reflective of the very different goals we have with these 2 studies, right? The APOLLO-B study was intended to, as rapidly as possible, try to demonstrate the benefits of RNAi therapeutics in this disease by looking at functional status and quality of life. HELIOS B, on the other hand, is twice as large and 3x as long. It's designed to look at outcomes. And so these subtle differences in these inclusion/exclusion criteria are simply reflective of those very, very different goals of the 2 studies, but the population of patients we brought into these 2 studies really is very similar and again, intended in both cases, just to be reflective of the population with this disease around the world.

Thanks, John.

Hi I am [ Zi Shu ] from Berenberg. Maybe the first one, just look at the subgroups if you compare the 6-minute walk and the first time to CV event. If you look at the tafamidis baseline versus no tafamidis at the baseline, it looks like 6-minute walk shows there seems no benefit in the tafamidis baseline patients. I know it's a small group. But then on the first time to event, it shows the opposite. Maybe you can help me understand the kind of -- how should we reconcile that? That's the first question.

Yes. I mean, maybe I can start and then Dr. Mauer, you might like to build on it. I think actually you're pointing out the challenges of actually interpreting small subgroups which is this is -- again, I think as Dr. Mauer said, these are -- this is -- in the overall scheme of heart failure studies, it's a pretty modestly sized study. The subgroups are even then smaller than that. And so -- and then you've got endpoints, which have an inherent amount of variability in them, particularly 6-minute walk test or low event rates and in a short study for outcome variables. And so I think you're just seeing noise around that that's creating these patterns. And I think I would encourage you to maybe take a step back and look at the totality of it. And if you look at those forest plots for KCCQ next to 6-minute walk test as Dr. Mauer presented, you will kind of see it just a holistic picture that really there's a very strong consistency of benefit across all those potential subgroups, all those subgroups that are listed but I don't know Dr. Mauer, if anything you'd add?

Yes. I agree. I mean I think you're just crazy to start getting involved in looking at all these little individuals that I can counter by saying, yes, the point estimate for the 6-minute hallwalk the tafamidis group was slightly different, but the confidence intervals are usually wide, and we all know how to interpret statistics, which means that you can't conclude one group is different than the other. I don't know, my grandmother used to say, you're getting involved in mishegoss, so I just wouldn't go there. I just urge you, if that's what you want to do, have a good time, but you can't convince anyone one way or the other, I think you have to look at the totality of the evidence. And basically, on the primary endpoint, the key secondary, all the exploratory endpoints. This drug is effective, right? And what I hope personally is that it will be not only as effective as it is now, but I think that probably and fortunately, over time, it's going to become more and more effective. And I made this point by taking care of these patients. This idea that we're going to reverse amyloid and make everything all better is, I think, selling patients [ a bill of goods ] and early treatment is really important. We've learned that. We try to identify people a lot earlier. That's one of the successes, and we try to institute therapy. So my concern, I guess, is about the effect size that people start talking about, it's kind of small. It's going to accrue a lot over years. And these people are going to be waiting for a few years to get access to the next level therapy. It could be a dramatic change in how they're doing. This is not a slowly progressive disease. This is a rapidly progressive disorder. So those are some thoughts.

Thanks, Matt.

Maybe just coming to your last point, Dr. Mauer. On the reversal of disease, obviously, within 12 months, we didn't see that. I guess have you -- do you have any data, imaging data to show any remodeling of the heart that can maybe in a few years' time that those patients might have a reversal of disease. Any thoughts on that?

I personally don't have data, but I would just point out, there are other amyloid diseases, not the focus of today's discussion, but in light chain amyloidosis that we've been treating with anti-plasma cell therapy for a long, long, long time. The name of the game is, I can't do anything to the heart until the hematologists get the light chains back to normal. And when they do that, -- there are many, many, many papers demonstrating that patients' hearts not quickly, but over time, can get better. And there's some dramatic examples that have been described and delineated. I mean it's public knowledge, but Isabelle Lousada who runs the Amyloid Research Consortium had multisystem ALM [ 25 ] years ago. And 20 years later, we barely can find it. So that's the hope is the body does have a natural mechanism to remove amyloid. And as Dr. Hanna was saying at tonight's meeting, we hope that by shifting the balance, getting rid of the precursor protein that's causing the problem, we'll put things in a favorable light. At the same time, I spend a lot of my time at the patient's bedside and with them in the office, and you have to set expectations clearly because it's not a short disease. It's a long haul. Thankfully, people are living a lot longer, thanks to all the efforts by all these companies, which is great.

Maybe just a quick follow-up on my last question. On the Class III, I know it's only 20-27 patients...

We make a rule that anyone under 100, we're not going to talk about anymore tonight. -- all the tafamidis patients.

I just want to confirm, those were the -- those are patients on baseline tafamidis or they are a mix of...

The Class III patients you're talking about. I don't know the details. That level of detail I don't have.

No idea. I'm not sure, but there we'd be getting in some...

Yes. I mean, like they are even smaller subgroups. I think with the point I would come back to you as is when you look at a small subgroup, then you'd have to redo the baseline characteristics in the subgroup and there are so many other things that drive stuff. -- example. In this disease, Black race is tightly tied to the V122I mutation, and those people do much worse. So you can interpret a lot of things from just race, but it's tied to the genotype and so forth. More often, patients with advanced disease have a genotype that's positive. So there are a lot of confounders in all of this and the subgroups. Overall, the groups are well matched at baseline, but if you start looking at subgroups, I have no idea whether they're going to be so well matched.