BridgeBio Pharma, Inc. (BBIO) Earnings Call Transcript & Summary

And thank you very much, and welcome to ATTR CM part 2. With us, I have from BridgeBio, CEO, Neil Kumar. And I also have Ananth Sridhar.

If you guys, I guess, first tell us about your efforts in cardiomyopathy, what's going on with acoramidis as far as where it stands right now since you reported results in December.

Yes. Well, thanks for the time, David. We appreciate the opportunity to be here. So our efforts in ATTR cardiomyopathy stem back almost to the advent of Bridge, maybe 5, 5.5 years ago. And at that point in time, as you well know, the mechanism of disease has been well described, some 140-odd mutations that seem to destabilize the tetramer transthyretin. And if you graph thermodynamic stability versus pathogenicity or penetrance, it was a fairly straight line. So a very nice genotype, phenotype. And what existed at the time and still exists today was the appearance of a rescue mutation or a trans-allelic trans-suppressor, the so-called T119M, that was a super stabilizer of the destabilized tetramer that appeared to be protective, at least at that time, in the context of the V30M variant. And so what we did in collaboration with Isabelle Graef and Mamoun at Stanford and now University of Pacific, the 2 academic founders that we had the privilege of working with, was effectively to try to mimic that rescue mutation to design a super stabilizer of the tetramer, therefore targeting the disease at its source, preserving the tetramer whilst, at the same time, trying to ward off what we believe is the pathogenic effector in this disease, which are those unfolded monomers that go on to deposit in the heart or the peripheral nervous system. And I could talk about chemically how we did that, but we were able to, I think, design something that was a superior stabilizer to what had been seen before. We took that into the clinic into a Phase I, Phase II, which we announced that data a couple of years ago, and then into a Phase III, which effectively has a nested trial design where we read out the disappointing results of our Part A 6-minute walk distance 12-month readout sometime in late December. And we're ongoing in a Part B that we continue to be very optimistic on for reasons we can discuss that will read out some time first half of next year against the hard end points of mortality and CVH. So that's kind of acoramidis at a very high level, yes.

Now when you actually compare tafamidis with your therapy, what are functional differences in how you think -- number one, how it's administered but also what benefits you think yours could theoretically offer?

Yes. And again, we've got to play this out with the data in the clinic. So I think I would caveat all that by saying we're running the appropriate experiment to do so. Our trial is quite close to ATTR-ACT's trial. But the salient advantages for our drug versus tafamidis were effectively that we see more target. We're less bound up in albumin. Effectively -- we bind the target more effectively, especially with the superior KD2 basically binding to the second binding site. And then I think most importantly, we have a different binding mode so we glue the target together more tightly. And the way we do that is by exploiting more of an enthalpic binding mode versus what tafamidis has, which is a little bit more of an entropic binding mode. And what we're taking advantage of there is the induction of hydrogen bonding towards the bottom of the thyroxine-binding pocket, analogous again to what that T119M is doing when it super stabilizes the tetramer, almost 40-fold increase in terms of the activation energy going from tetramer to dimer in the sort of unfolding mechanism, as has been shown, I think, in multiple papers. So those are the 3 primary advantages: we're seeing more, we're binding better and we're gluing more tightly upon binding. And that means that we're about 2x better in terms of the stabilizer tafamidis, as has been shown in multiple experiments. It's about a 50% stabilizer at 80 mg. We're close to 95% or so. And that means we're about a 50% elevated serum TTR levels, about 40% actually as we saw in Part A. And so all of that, we hope, will ultimately lead to, again, less of that monomer depositing in the heart over time, which feels like it could lead to substantial benefit to patients.

Before we jump into the specifics of the trial, just curious to hear your thoughts on -- obviously, it depends a lot on ONPATTRO, what they do in their study for APOLLO-B. How do the stabilizers and silencers fit in the market together?

Yes. Well, I think we're all acting on the top end of the path of mechanism here, which is to prevent the deposition of these toxic monomers in the heart in the case of cardiomyopathy. So in that sense, we're doing something very similar. And as you see in polyneuropathy, a similar amount of knockdown is about tantamount to a similar amount of stabilization. You can see that if you compare diflunisal, which is about 70% stabilizers results; versus inotersen, which is about a 70% mean/max knockdown. And that makes sense if you do some fairly simple molecular modeling. You're basically doing the same thing and in effect, trying to ward off again the amount of toxic monomer that's being produced. The advantage, I believe, in using a small molecule approach is that you preserve what's otherwise a very important protein in the tetramer. And it should be said that there's no human that's haploinsufficient or null for TTR. Elevated levels of TTR over time seem to be cardio-protective and correlated at least with a longer life. There's no species that I know of that's haploinsufficient or null for TTR. And the protein is 1 of the top 50 most abundant proteins in our body, and it's turning over once every 2 or 3 days. So that's a pretty high energetic cost that it wasn't useful. So all things being equal, I think one should be able to establish the same efficacy, meaning you can limit the amount of monomer as productively using a stabilizer versus a silencer or a knockdown. And I think secondly, you're preserving what could otherwise be an important protein. So hopefully, you do it more safely. Again, with the caveat that we do all this -- yes.

Got it. Now we look at the study itself. Clearly, the secondary end points looked positive. It missed on the primary end point. I know you started drilling down into the different class levels between yours, tafamidis' trial; frankly, between your experimental arm and your placebo arm as well. What dynamics do you think caused the 6-minute walk test to perhaps fall short? Could it have been something in the nature of the screening of patients and the types of patients in the trial?

Yes. I mean there was -- I think we've gone through this in pretty great detail. We probably don't have time to do all of the various hypotheses, but it didn't seem to be discrepancy in the class of patients. We were modestly more healthy than ATTR-ACT in terms of the NYHA mix. But even if you had corrected for that based on our results, you wouldn't have had an experiment that hit or a placebo that declined. And that was the issue, as you well know. Placebo didn't decline, so everyone was effectively healthy. It wasn't -- we had about 10% variance. If we had 20% variance or closer to what ATTR-ACT had, it's -- that wouldn't have made a negative trial positive or a placebo that decline. We had 75% technetium scan versus 25% biopsy. There was no difference between those 2 populations in our trial. And so there's nothing immediately obvious that pops out of the baseline characteristics or the evolution over time. Well, I think we've commented earlier AE-driven death looked pretty much the same as it did in the ATTR-ACT trial, if you assume that AE-driven death is about half of all deaths, which is what they observed at 30 months. The declination on KCCQ or quality of life was pretty similar to what ATTR-ACT saw. The increase in BNP was pretty similar to what ATTR-ACT saw. So we did see a declining patient population against every other measure except for 6-minute walk. And so one wonders why. I think you're right to ask the question. And I think there's probably 2 competing hypotheses. One is, as you well know, 6-minute walk has a fairly large standard devs, almost twice the mean and -- as observed in ATTR-ACT, as observed in our data set. So it could be that we just deal an unlucky card, like an 8% likely number at the end of 12 months. The second is that there's been an evolution in standard of care. People use fewer rhythm control meds, they diurese more aggressively. And so there's -- there could have been that plus context and training bias, meaning -- well, I think it's obvious what that is. So those things could have been contributing. And I think we'll learn more from APOLLO-B. Four data sets that we designed our trial on prior to all suggested that the decline would be 40 to 60 meters over the course of 12 months, and I think ATTR-ACT's was like minus 56 or something like that on placebo. I think that there was a paper that was published after our trial kicked off out of the [ NAC ] that suggested minus 11 6-minute walk distance over 12. I think it was minus 11 in the white paper, right? So yes. So I mean we're doing experiments now also there that suggest that maybe you don't get as much of a decline but you still get the other factors headed in the wrong direction for patients, so there's still an opportunity to help them. But we'll have to see with greater data, especially coming near term from Alnylam.

You spoke to the random error of the measurement itself. Does that random error vary based on what class level the patients are? [ Would it be higher with ] patients of class 3 versus class 1?

I didn't -- I don't remember exactly what the standard deviation was in 3 versus 1, but I don't remember there being a striking difference. I don't know, Ananth, do you remember?

No, there's not a striking difference. I mean what we did see was that individuals with lower baseline 6-minute walk distance had higher variability, but it wasn't broken down by heart class.

Got it. Got it. Would you know the breakdown of mutant versus wild-type patients in the study?

Yes, it was 90-10. Yes.

Was there a different level of severity or different levels of performances in those 2 groups?

Yes. I mean you have to -- it's 10% so you have to take a little bit with grain of salt. But in the mutant population, the drug did flatline 6-minute walk and you saw a declination of almost minus 40 on placebo. So there was a significant effect there, and we did -- it was -- there was a decline on placebo. If you just assume that 20% of your patients were like that though, you still wouldn't have hit a p-value of any sort. We're obviously not going to run 100% [ aberrant ] trial. But yes and I still -- I don't know what to make of that because the stats are so small.

Got it. Got it. And was there, I guess, a variation in the ventricular thickness -- wall thickness for patients? I know the minimum was, I believe, 12 millimeters. Were most patients close to that? Were -- was it a broad range?

I don't remember exactly what the range was above 12 in terms of left ventricular wall thickness. I don't know, Ananth, you might just mention what the inclusion criteria were broadly. If you...

Yes. So it's a good question. I think the minimum was a wall thickness of 12 millimeters. I mean there's not really strong evidence correlating more -- bigger ventricles with outcomes, at least from the evidence I'm familiar with. But our general eligibility criteria mirror what was studied in the ATTR-ACT study and most of the main components of eligibility from the patient population perspective. I think the key differences to call out is our -- an increase in the baseline 6-minute walk distance. We raised the minimum to 150 meters from 100, which was studied in the ATTR-ACT study. This was after a review of the data shared from the -- in the NDA review, where we saw individuals below 150 meters at baseline deteriorating quite quickly regardless of treatment allocation. And then we increased the upper limit on the NT-proBNP to 8,000 picograms per ml.

And I guess my one last question before we start looking at the what's next with the study, on the December results is in the diagnostic process itself, certainly, there can sometimes be a challenge. And when you see patients with amyloid, of -- it could be amyloid of another cause, something to confound the results. Do you feel very comfortable with the study as far as -- that all the patients, in fact, their amyloid was TTR?

Yes. Ananth, do you want to...

Yes. So we did follow the guidelines for diagnostic criteria in the study. So like Neil mentioned, it was about 75% of individuals diagnosed via scintigraphy, 25% via myocardial biopsy, all patients underwent serum light chain screening to rule out AL amyloidosis. And we also had a diagnostic confirmation committee to confirm the image read for each of the individuals screened into the study.

Excellent. Now of course, we keep talking about the 6-minute walk test, which is an unfortunate result because all the secondary endpoints were in favor of the drug. And the DSMB has decided not to make any modifications in the study as it proceeds to 30 months. So to this point, I guess, what do you know about -- I mean I don't know what you've seen to this point on outcomes. Obviously, the DSMB has seen something enough that makes them feel comfortable. And how are you thinking about going into the 30-month portion of this trial?

Yes. I think it boils down to 2 questions really. Number one is, to your point, do we believe the drug was safe and biochemically active in a manner that's supportive of having an impact on the final hard endpoints. And I think from a wide variety -- well, on every other measure, as you just talked about, it seemed like we were having an impact. I think most importantly for us, seeing that biochemical impact on serum TTR, which is quite a bit higher than I thought it would be given that we were 90% wild type; as well as seeing the BMP effect and BMP being one of the better vectors of mortality. I think -- and then for small numbers, again, AE-driven death, to see that safety there and an imbalance in terms of an advantage for treatment, I think all of that stuff came together to give us quite a bit of optimism. But the second question that one needs to consider is somehow the standard of care or something like that, such that we're not going to have enough events so that we can see an effect of the drug, right? Is there enough of the decline on the baseline of what we care about going forward that you could see some impact? And that's really what we took a look at in mid-first quarter of this year in a blinded manner to look at all-cause mortality to see that if you make a few corrections to ATTR-ACT, that were well in line both with what they saw and what other natural histories have seen, unlike what we saw on 6-minute walk. But I think, suffice it to say, we were comfortable with the numbers that we are seeing, which were well in excess of either a healthy cohort that will be 77 mean age and in line with what we would have expected to see in a TTR cardiomyopathy patient population. So you put those 2 things together, we continue to believe that Part B has a good shot of success.

Yes. The interesting thing is the 6-minute walk test, that primary endpoint, but -- in real-world clinical practice, that's not going to be the driver of use by any stretch. The doctor is not going to say how are you walking today versus last day -- the last time. They're going to make the decision purely based on outcomes. Does it really make a difference in the long haul for any of these drugs if they can achieve on outcomes without achieving on 6-minute walk test, given the nature of the endpoint?

Yes. Well -- so let me make a comment as to why we included the endpoint because we obviously didn't include it in our Phase II. And you're right to say that hierarchically, people are going to care a lot more about mortality and CVH outcomes. I think by and large, everyone likes to connect the dots, right, in the space. And I think that mortality is going to be -- hazard ratio is going to be the #1 thing, according to our research, that's going to ultimately drive scripts and share in this marketplace. But you'd like to see that if you had an impact on mortality and impact on hospitalization, that then ported to other things that you cared about, like BMP levels and serum TTR on the biochemical side and like, hopefully, KCCQ and 6-minute walk on the outcome side. And my best vision is as you see deterioration over 30 months in 6-minute walk, you will see the drug separate from the placebo arm. It's just the 12-month endpoint, which we really derived from the ATTR-ACT trial and we really put in place so that we could run a clean placebo up through 12 months, that I think may be a result that -- clearly, a placebo result that we could not replicate and may not be repeatable just based on where we're at in terms of caring for these patients today.

And we're kind of getting up close to the end here. Tafamidis comes off patent, goes generic 2026. Assuming all goes well in first half of next year, you get launched in 2024. How does tafamidis going generic impact acoramidis?

Well, I expect they'll get PTE at least. So you're talking about early -- 2031 or 2032 somewhere in that realm. And then they have quite a bit of IP protection in and around the formulation here. The meglumine salt is nontrivial to make, as you know. So there may be ways that they can extend it past then, but certainly not going to be '26 when they go generic in the U.S. market. Look, I think you can look back at any other cardiovascular category. My suspicion is if we're right on the path of mechanism of disease, we could get a 15% relative risk reduction versus them on mortality. And if that's the case, then I think it's not like everyone moved from the second generation of statin to a generic Pravachol or whatever generic. People are going to use the thing that has a superior effect or impact on mortality. So I think that could protect us from genericization in the 2030s time frame. And I think we'll be looking to do other studies should Part B be positive to demonstrate the superiority of our compound, both in specific variance settings as well as in earlier settings, almost like primary prevention like setting. So those are all the key things that hopefully will help to build a fence around our drug, if indeed, it is considerably better than Taf 80.

Got it. And with that, we've come up to the end of our time. Thank you so much for being here today. Appreciate it, learned a lot, and look forward to chatting again soon.

All right. Cool. Thanks for the opportunity. See you.