Awesome. Well, excited to be here with the NewAmsterdam team. Very exciting time for you guys, a lot going on. We have Michael Davidson, CEO, and Matthew Philippe, Head of IR and Finance. Maybe we can just jump right in. There is obviously a lot to discuss. I think investors know this story well, but maybe before we get into PREVAIL, we can just kind of talk about some of the takeaways from AHA this weekend and just sort of any implications for obicetrapib.
Sure. Yeah, thank you. Thanks, everyone, for coming. Just got back from American Heart in my hometown of Chicago, and it was obviously the two big studies was HORIZON and ZEUS.
A lot of very important learnings. ZEUS, of course, that was my drug, Corvidia, sold to Novo, ziltivekimab. I still think that a lot more to be learned from that trial, but it ended up being a very sick population. I think more to come. I do not think the inflammation story is over by any means. Even IL-6, I think there might more come out of that as time comes on. But I think what is a learning there is that we have learned this already in the LDL field, that when you go too sick in a population, it is too late, and you cannot have a benefit.
It was designed in some ways for that because we know that statins do not help this population because they already have heart failure, kidney disease, and we thought inflammation would be a driving effect that we could mitigate with IL-6. I think we will learn more, but I think there was too many patients were already too late.
I hope that more comes out of that. ARTEMIS is ongoing, less sick population, which might see a benefit. That is ZEUS, and I think it factors into the overall pessimism of CVOTs that we are now facing in the field, which is, that is the life we lead in trying to develop new therapies, but we are not giving up, of course, and remaining optimistic about the whole anti-inflammatory approach to reducing residual risk. HORIZON is, I think, another important learning, and it has a lot of read-through to us, but I do not think how people think about it the way I think about it as a clinician.
Up until HORIZON, because I see a lot of patients in my University of Chicago clinic that have high Lp(a), and what we have been telling them for a few years now is, "If we get your LDL low and your CRP, your inflammation marker, low, the risk is substantially lower.
You have high Lp(a), let us get your LDL down below 50. Let us get your CRP as low as we can. That is usually weight loss because visceral fat is the most inflammatory contribution there. If we do those two things, we can make a big difference on your events. HORIZON confirms that decision because overall, there was very low risk if you had a high Lp(a) and you also had a low LDL and a low CRP. In fact, the numbers are pretty astounding, actually. If you had a high LDL above 100 in HORIZON versus a low LDL below 100, the risk was something like 10%-18%, more than 50% increased risk for having a high LDL versus a low LDL or lowish LDL. That is the PREVAIL study. The PREVAIL study is a high LDL over 100 is our median.
Median is over 100, and we have high Lp(a) in a subset of patients, but it is not a high Lp(a) trial. I think one of the takeaways that we have heard from investors is, "Oh, now that Lp(a) is dead, the PREVAIL trial, that added benefit in PREVAIL is no longer going to happen." Any differentiation that obicetrapib had from other LDL drugs is also diminished because the Lp(a) lowering benefit is no longer part of that overall benefit that obicetrapib delivers to patients.
When you look at HORIZON in those contexts of the subgroups, you see that in that upper LDL population, which is like PREVAIL, you saw not only much higher risk, but also a 0.83 hazard ratio for pelacarsen, which suggests that there might be some benefit when LDL is high for lowering Lp(a), and that's what the PREVAIL study population is all about. I wouldn't write off that there might be something extra about Lp(a) lowering that obicetrapib provides that could contribute to the PREVAIL benefit. I think obviously, there's a lot more to that. Obicetrapib raises HDL 150%. It lowers the risk of diabetes. It lowers all the small particles by 90%. Of course, the Alzheimer's benefit. We have many differentiating features.
In addition to the fact it's a small little pill that you don't need to worry about food effects and no side effects more than placebo. In combination with ezetimibe, we're in that 50% plus range, which is comparable to any other PCSK9 inhibitor for that matter. We think HORIZON does, in fact, give us more optimism about obicetrapib having an extra benefit in PREVAIL.
The other thing, of course, it provides is that new- If there is no benefit for Lp(a) lowering for secondary prevention, we think primary prevention is another. That came up quite a bit in the discussion, that primary prevention for Lp(a) lowering is where the risk is really the most obvious.
What's interesting to me about HORIZON is the average age was 59 in that trial. These are patients post-MI that had high Lp(a). In a typical CVOT of the same type of patients, the average age is over 65. These are younger patients having an event, which suggests that for Lp(a), the biggest issue, you get more premature disease. It leads to earlier heart disease. Primary prevention really should be the focus.
And we have other studies ongoing that will look at primary prevention, but obicetrapib could become the main primary prevention-type drug for these patients because, again, talking to patients, they don't want a statin, which raises Lp(a). They'd rather have a drug like obicetrapib that lowers LDL and lowers Lp(a), if they have high Lp(a). So we think it really helps differentiate us even more so than any other lipid drug going into the market. And I think the main point is that differentiation, we believe, is still there for Lp(a) benefits. We think PREVAIL, HORIZON does give us, I think, insights about that PREVAIL higher LDL, higher Lp(a) group could have an added benefit.
And the last point I want to make is that because all these other Lp(a) drugs have gone away, and ZEUS has really diminished the inflammatory kind of drugs out there, we become the next big drug out there for cardiovascular, assuming PREVAIL works. And so that's going to create a lot more, I think, excitement for our drug, once it's approved.
And we'll see what other companies need to do to fill in their gaps in their pipelines, and we become an obvious option for those-
Yeah
for those companies.
Yeah.
That is a long answer to your question.
No.
I want to get to the-
It is thorough.
I think to get to the HORIZON messaging, I think, is for us, we have been saying all along, it is a win-win for us-
because of those other benefits that we believe the drug has. If we do not see a benefit in HORIZON, we still become the first Lp(a) lowering drug, quote, "out there" for the patients.
Not to make the answer longer on HORIZON, but I think one of the things that also helped us understand is a concern that a lot of investors have, and that is kind of been echoed in the industry, is that standard of care is bringing down event rates. I think what we clearly see, and we have been saying this for a while, is we already have the VESALIUS data out. We have seen the more contemporary studies, and placebo event rates have been pretty consistent for the last decade.
I think now, obviously, remember, HORIZON was running the same time as VESALIUS, but to actually look at HORIZON and say, "The reason your event rates were so low is because your LDL-C risk was so low." If you actually look at their event rates in that above 100 group, it is exactly what you would predict. As Michael said, it was 18% over the study. I think for us, that is another validation that placebo event rates really have not changed. The underlying risk has not changed.
once you adjust for LDL-C. I think for us, that creates more optimism as we head into PREVAIL on the sense that our trend that we are seeing in our blinded data is now even more or less likely to be driven by standard of care placebo changes.
Yeah. I was going to say, I feel like that is beyond that subgroup, which is really interesting. I feel like it is just another contemporary confirmatory kind of dynamic of LDL-C drives MACE.
Yeah.
I think, to your point, there has been so much focus on HORIZON, these benefits beyond just LDL-C for obicetrapib. But in the context of LDL-C alone, we obviously know it is quantitatively predictive, and it has been very quantitatively predictive. Just based on sort of the baselines, the length of follow-up here, what should the LDL-C math kind of set as a floor for MACE in theory for obicetrapib?
Well, we have said that we look at the LDL, non-HDL, ApoB lowering in BROADWAY and how that models for PREVAIL, because again, with our experience of doing trials, we wanted to do a couple things. We wanted to have a high baseline LDL.
We wanted to have it go long enough to see a benefit, and also, we wanted to have a preliminary phase III trial, lipid trial, that would match as best we can the outcome study. So we could use the phase III BROADWAY trial as a good, not just de-risker, but model our event rates.
and model our LDL, non-HDL, ApoB lowering in our outcome study. We are very fortunate in that regard because the populations match very closely.
If you look at the LDL, non-HDL, ApoB lowering in BROADWAY, we can then apply that same kind of reductions in PREVAIL, we come up with the well-established correlations between those three parameters, best being LDL, because LDL has the most data.
We are in that close to 20% relative risk reduction, just in LDL alone. Same for non-HDL. HDL a little bit better, like 22%, and ApoB is like 18%.
Yeah.
We are not far off from any of those three parameters.
Yeah.
18%-22%. Rather use ApoB versus non-HDL cholesterol.
Yeah.
I mean, I think that's important as well. It's like, you can look at it several different ways. You can look at the CTT-type analysis and say, "What does that predict?
You could say, "Okay, now we have also the REVEAL analysis. Where does REVEAL predict we should be?" It's also around that 18%-22%. We have BROADWAY data, obviously, that showed 21%, and we also have our blinded event rate trends that we've been monitoring as well. Each of those different buckets, independent of each other, are all pointing at the same direction. I think for us, that's the optimism. The last point of optimism, I think is probably the most valuable, is Michael and John.
Yeah.
They've been part of CVOTs for a very long time, and what they're seeing excites them, and I think that comes across. For us, everything's kind of pointing in the right direction, that we're in a good setup heading into the interim look.
Yeah. No, that totally makes sense. I'm glad you brought up ApoB because I feel like just again, in the context of all these CVOTs failing, people are like, this ApoB dynamic comes back up where, I think on a relative basis, maybe it's a little bit lower than what you've seen in terms of LDL-C. Maybe just sort of, I mean, you're kind of still saying it gets to 18%, maybe just kind of like talk about some of the nuances with ApoB versus LDL-C and-
Yeah
again, sort of the math maybe that kind of gets you to that 18% on ApoB.
Yeah, I mean, based on the statin trials, every 10 milligram per deciliter drop in ApoB leads to a 12% relative risk reduction.
Our baseline ApoB is roughly 100, 18% reduction on the actual ApoB lower, because I think our paper had 13%, but that was the imputed number based on retained dropouts. The real difference was 18%. Some people have used that number a little bit off. So 18 milligrams per deciliter, you multiply 10, you get to that 18%, basically, relative risk reduction, even closer to 20 sometimes, depending on how you look at it. But that's where ApoB falls in.
Yeah.
Again, but I think people look at ApoB, and it does have kind of a mythical effects to it because I am an ApoB advocate too for its predicting what is called discordant risk. You have high LDL and high ApoB and low LDL, ApoB outperforms LDL. But as a risk predictor, LDL always is where all the data really is.
Out of 40 different trials with statins and other lipid drugs, the one that most closely matches the line is LDL. And then for CETP inhibitor, what they found with REVEAL was non-HDL most closely tracked-
the event reduction. That is why they used non-HDL on the same CTT line. So we think, if anything, because REVEAL was a CETP inhibitor and we are a CETP inhibitor, we think non-HDL is the more likely predictor of benefit in PREVAIL.
Yeah. Makes sense. REVEAL, I think, is really interesting because you can do sort of all this math around these lipids, but that's like a large empirical data set that similar number of events as maybe in PREVAIL, long follow-up, and that drug was half as effective on LDL-C cholesterol, and it still showed a 17% MACE. Is that like the right take? Is it that sort of simple of like, this is a very representative data point and our drug is twice as effective?
Right. I think it is the best. In other words, if it wasn't for REVEAL, we wouldn't be here today.
I mean, we looked at that study and said, "This is a validation that what we're doing is just find a better drug and do a better study, actually replicate the upper tertile of REVEAL with a better drug, and we should be." Again, we don't like ACCELERATE. They never had a prerequisite study.
Yeah.
Ziltivekimab just didn't have another. Here we actually have a previous study done, and the population overlap between REVEAL and PREVAIL is like over 90, almost 90%-
in that upper tertile.
Yeah.
We cannot ask for, again, a better comparator.
Yeah.
I know David Ricks tweeted out yesterday. I don't know if people saw it or not.
Yeah. A couple people were like, "What do you think of this?
Yeah, a lot of investors say that CETP is a properly druggable target, not working, but he refers to evacetrapib.
Yeah
Which he got burned on. I understand, that's the They neglect to talk about the anacetrapib.
Yeah
Trial in REVEAL.
Yeah
They obviously heard that comment.
Yeah. I mean, there's obviously so much focus on the interim, but it's an approved drug in Europe.
Yeah.
And yeah, just like on that Lilly thing, just because it has come up a few times because you can look at the upper tertile in the Lilly study as well, much shorter sort of follow-up period in terms of why that didn't separate. It's a much shorter follow-up period. Is that just like the simplistic conclusion? It just wasn't followed long enough, or?
That is the most silly, because look at the amount of LDL lowering was way overstated by that direct assay.
We know it was a wrong assay. We have ezetimibe, and this was an acute coronary syndrome trial. Then we know the separation does take longer. If you look at IMPROVE-IT with ezetimibe, which almost had identical LDL lowering, at two years, the lines were superimposable. They had not separated yet.
ACCELERATE, if you look at that population, not separating in two years, it fits exactly there. We think if they would have gone longer, it would have had a beneficial effect.
I think Lilly was in a difficult position at the time, and they wanted to obviously minimize R&D costs, and that was something that just decided to do. Also, there were other challenges with the drug. It was very hard to manufacture, very expensive. So they just made some business decisions-
Yeah
based on that.
Yeah. No, that makes sense. I think, like you said, Matt, you can kind of do the math on sort of the lipids. We have REVEAL, the event rate dynamics that are actually happening in your study, and definitely want to kind of talk about event rates. But first, like you said, I think that all kind of triangulates around a compelling number, and typically what people are kind of referring to is MACE-4. Obviously, MACE-3 also kind of has to hit as well. What have we kind of seen in terms of how MACE-4 and MACE-3 track in terms of effect size across other CVOTs? I am sure you have heard this, but I think that the one sort of pushback on BROADWAY is the effect size was driven by revascs.
Right.
So, sort of what is your confidence that it will also track in terms of MACE-3?
Right. There was roughly. So the event rate. So in MACE-3, there was no overall benefit, although non-fatal MI and CHD death were in the right direction. Stroke was the opposite.
They are a very small number, only 19 stroke cases.
It is hard to if you get just a little bit of wrong direction offset. The 3-point MACE is roughly 2.5%, and the 4-point MACE was roughly 5% in BROADWAY.
As we showed already, PREVAIL, first year, the event rates track almost exactly what we saw in BROADWAY.
For not just 3-point or 4-point, but also non-fatal MI, stroke, total revasc, urgent revasc. All these events were very close to BROADWAY, which is what we were hoping for, that BROADWAY would be a really good de-risker. That is year one. And in year one, as we know, in BROADWAY, the 21% risk reduction kind of split the placebo active. That is where PREVAIL kind of splits the blend, it splits the two Kaplan-Meier curve-
lines, which again is supportive of that. We are seeing something similar in PREVAIL as we saw in BROADWAY. But then year over year, we then track year two, year three, and what we are seeing, of course, is MACE-4 rates annualized, and MACE-3 rates are going down.
They keep going down. With knowing that, the going down rates, we can apply the placebo rate from BROADWAY or the first year PREVAIL rates as placebo rates, then we can impute what our benefit will be at the interim.
We come up with an optimistic stop at the interim.
We can also say, let's say somehow we're off on the placebo rates for PREVAIL. We've looked at all these other trials, CLEAR Outcomes, VESALIUS-CV, now HORIZON, and we're finding that the 3-point in particular is a really pretty stable event rates across the trials when you adjust for LDL levels and things like that, diabetes, age. We feel good that even if we use any of these other placebo rates in our imputation in PREVAIL, we still look good.
We not just took our own event rates in BROADWAY, we also looked at a lot of other trials as well to try to estimate what a placebo rate should be in PREVAIL. When we apply these other trials, including now HORIZON, we still look up, we're still optimistic about the interim-
Yeah
being positive.
I think in terms of like the MACE-3, MACE-4 component, as Michael mentioned, although MACE-3 on its own was not separating, we are already seeing very compelling trends in MI and CHD death, actually all-cause mortality as well.
It was only stroke that was kind of going the wrong way, which is expected. If you look across all the CTT studies, if you look across like the PCSK9, LDLC studies, they also had stroke going the wrong way. The CTT studies, stroke in year one is like a 0.96 hazard ratio. It is a 0.77 in year two.
PCSK9s, it was going the wrong way, was a big driver of benefit in the outcome studies. It just takes longer for stroke to separate. We also are seeing that with the blinded data. We are seeing very compelling stroke trends that are continuing to show this same move. I would say the other data point to really look at is, in most studies, MACE-4 separates first.
Stroke separates later. By the second year or end of study, MACE-3 usually is better than MACE-4.
What that means is MACE-3 has to accelerate in that second year to catch up and pass it.
Yeah.
Actually, ODYSSEY was the only study, contemporary study where MACE-3 was the same as MACE-4. MACE-3 in most studies is better than MACE-4.
Yeah.
We're also seeing those trends in the relationship between MACE-3 and MACE-4 that are kind of saying, "We're seeing this catch-up phase for MACE-3." I think-
Yeah
All those kind of point us in the right direction. Obviously, it's blinded data. You can never be too certain with that. I think, fortunately, it's at least a little helpful.
Yeah, no, definitely. I think VESALIUS-CV has outperformed, like the very recent LDL-C study outperformed, MACE-3 outperformed MACE-4. I guess like, you're based on all those data kind of confident around replicating a BROADWAY-like effect size. We know MACE-3 and MACE-4 track pretty consistently. As it relates to the interim itself and sort of the powering, I know you guys haven't been like too specific, but in terms of the number of events, you're very confident in there being, with inter effect sizes like that you will sort of hit the P value threshold and stop.
Yeah, we're pretty confident. What we've said is, the thresholds we set for the trigger for the interim, we're well powered at the interim if we show a BROADWAY-like effect size.
Yeah
For the interim to hit the stopping criteria, the alpha levels that we put out there.
Now obviously, do not confuse powering with minimum detectable difference.
Yeah.
There is room around that. But I think that is why we chose the event number we did.
Yeah.
Because it allows us to hit on what we think the effect size is, and potentially even lower.
Yeah. And I guess, so in that context, is your base case that this will stop at the interim?
I think the data is definitely giving us cause for optimism.
Yeah.
But I think we also are very well aware that there is just random variability that can happen in clinical trials that could offset something. That is why we said, "Let's go a year longer," because that removes any of that random chance from the potential outcome.
I think that is the ultimate goal for us, and what we have seen is, in the market and the way guidelines are written and the way physicians use therapy is, the magnitude of benefit is not as relevant as having a benefit.
It is a zero or a one. If you have the benefit, then you get the usage.
I think for us, that is what allows us to go the extra year longer, because that kind of removes random chance off, random variability, of just getting unlucky at an interim time point that we do not have at the end of the study. Because if you look at all studies, it is never that VESALIUS hit 19% MACE score at year two, and it was 19% at every time point you looked as it went forward.
There were periods when it was actually 25%. There were periods where it was 17%, 16%.
Yeah.
I think it is important to realize, you can just have that random movement, and we want to take that off the table. That also tells you, there is no difference between a 15% or a 16%, or 18% and 20%. Those numbers could just have to do with when you stop the study.
Yeah.
We want to make sure that we are well set up heading into that. All the data we have says that we are very well powered for the interim.
But we've given ourselves the opportunity, if something does go off for random chance, that we kind of remove that by the end.
Yeah. That's what I was going to say. I feel like the skeptical view is if the interim doesn't stop, that means the effect size is 15% or less than 15%. Is that not the right conclusion?
Well, but it could get better over time.
That's one thing. That's one factor. But yeah, I think, as Matt has been saying, you can't assume anything for sure. But I would say the most important thing for us is a successful trial.
Yeah.
As you can see what Amgen's doing with VESALIUS, they had 19% three-point MACE, sorry, four-point MACE as their primary endpoint. They're tallying a 35% fatal/non-fatal MI rate in their promotion. The relative risk reduction part to us is people making too much of a big deal about it. It's all about whether you have a positive. All the KOLs care about is a positive trial or not.
For us, we want to make sure that we maximize the chance of a successful trial.
Yeah.
That's why going at the end to the one-year mark, with the much more events, we don't want to take any chances. We have one last chance with CETP-
Yeah.
to get it right, and we're not going to take any chances. The interim, though, gives us that chance to get on the market a year early.
Yeah.
I know it's a big factor in our stock, and people have written us off as unlikely to hit the interim, and the risk that, why are we taking this risk? But for us, we feel it's a risk worth taking, in light of what we're seeing in our blinded data and what we saw in BROADWAY and what we're seeing other clinical trial placebo rates and how that adjusts and REVEAL and so forth. We think it's a risk worth taking to get on the market a year early and then get ahead of some of the competition and really be in a good place with the momentum going now with the market growing so rapidly.
We want to be part of the solution as fast as possible.
Yeah. No, totally. There's a million other questions I could ask, but we're obviously up on time. Maybe just to kind of tie the bow, so disclosure in 1Q, and it's just going to be the initial sort of press release is just going to be it stops or it continues. We're not going to get any-
Yeah, unfortunately, that's all we'll know as well.
Yeah.
The recommendation from the DSMB, we are still blinded. They will tell us their recommendation. They will not tell us why they made the recommendation.
Yeah.
We will have to formally close down the study and lock the database, and then we will finally become unblinded to it.
Yeah.
We expect their recommendation in 1Q, and then hopefully shortly thereafter. What we say is, if their recommendation is to stop, whatever number they saw will be a highly clinically relevant number.
Yeah. Makes sense. Awesome. Great. Well, thanks, guys, for joining.
Thank you.
Thank you.