第188期｜2026年7月10日

You're listening to Biotech Hangout, a live and unedited weekly discussion of all the latest news. We try to do all the latest news anyway in our industry with a group of biotech leaders and experts. I'm Sam Fazeli, and my co-hosts today are Josh Schimmer, Eric Schmidt, Paul Matteis, and special guest Matt Herper from STAT. For more information about our hosts and guests, speakers, or to listen to the most recent episode, please go to biotech hangouts.com.

So, a bit of a bubbly day today for biotech. I've been looking forward to celebrating one of the highs of the XBI, but maybe we don't care too much about the XBI. The index is down, though. There are some massive share-price moves, and I can't get my head around it. I don't know why.

Some colleagues pinned it on the “ceasefire is over” comment, but oil prices aren't moving, so I don't know what's going on. Maybe it's just a little bit of profit-taking. Let's start with that quickly and see if anyone—Josh, who is not muted—can start and tell us what you think. Then, folks, jump in before we move on to the regulatory policy stuff.

Well, the XBI is still up 30% year to date, nearly 80% year over year. So, a little pullback from time to time is not any cause for concern. We'll see what the future brings. Maybe things were arguably getting a little frothy with some of the M&A speculation, which is always a little bit of a dangerous place to be.

On the other hand, industry fundamentals are better than they've ever been.

Yeah, agreed. Eric, anything to add?

Oh, we can't hear you. Okay, if anybody else wants to add anything, you can jump in as I move on to the next subject. You can push into this if you want.

Regulatory-wise, there's quite a bit going on. We got Matt on today to talk about all the wonderful stuff he's been writing on and the interviews he's done on AI. Let's start with something all of us could talk about. I think it was Eric who brought this up: the CRL walkback—the complete response letters that the FDA, under Dr. Marty Makary, had said it was going to publish, and now there seems to be talk of retrenching from that position.

There's also been some talk about Rick Pazdur coming back, so that's going to be interesting to see. Let's start with the CRL walkback while Eric waits to join us. Let's go to Paul and Matt to get your thoughts on that. Matt, you can tell us a little bit more about Pazdur, given that you spoke to him recently.

All right. Am I going first, or Paul?

Go for it. Go.

You go ahead, Matt.

On the CRL walkback, I wrote a lot about Dr. Makary. One of the big issues throughout his tenure was not really worrying about legality or regulation before making a decision, and that can result in things being walked back.

On the CRL transparency, which I thought was one of the best things he did, there's a reason it hasn't happened before, and it seems to have been challenged in court. I spoke with Rick at J.P. Morgan, onstage, shortly after he left the FDA, and I spoke with him again at ASCO, at our event there. Both of those were long conversations, and we wrote about them.

I would just caution that I've known there have been discussions and that he is definitely in the mix. When he was at ASCO, other people on the panel were suggesting him as a potential commissioner. I think Dr. Pazdur probably would want a role where he has some substantial control if he were to go back to the FDA.

I also think we shouldn't underestimate that the Trump administration hasn't loved people who are critical of the administration and bringing them back into the fold. So, while it would settle a lot of nerves to have him in a high-up position, I have to imagine that those discussions are ongoing. I don't think that they're necessarily going to be the easiest circle to square.

We don't even know. I think they're going to have to decide on a commissioner first, before they decide who all those other people are, even though we have heard that Trump wants to announce all of this at once. It's great that they're considering him. I think they'd be crazy not to consider him.

I wouldn't say that just because he's being considered, we should all assume that we're going to have Rick Pazdur back at the FDA, making all sorts of fun noise in the way that we are accustomed.

Right. So, Eric, while you were rejoining, Matt said that one of the good things Dr. Makary did was make the CRLs public, but that there may have been some issues there—legality and so forth. I wanted to see if you wanted to chime in on that, and then get Paul's and Josh's thoughts on both Pazdur and the CRL walkback. Eric?

Yeah, thanks, Sam. Hopefully you guys can hear me now. I agree with Matt's comments that one of the very few things, maybe, that Dr. Makary had done when he was commissioner was become a little bit more transparent with regard to these CRLs.

We heard through various sources—and BioCentury did a great write-up on this this week—that almost immediately, a variety of law firms challenged the legality of that release, citing that this was a breach of proprietary confidential information on the part of the sponsors. In fact, multiple other prior commissioners had considered releasing the CRLs but ran into the same legal roadblock.

What's interesting to me, and I hadn't appreciated this, is that we haven't seen CRLs released since April of this year. It's been 3 or 4 months since the FDA has released a CRL. Apparently, that stoppage coincided with a citizen petition filed by Covington & Burling, a Washington law firm.

There may be a meaningful challenge underway right now that prevents investors from at least getting some of the transparency that I think many of us have relished. I don't know where we would be had we not seen, in gory detail, some of the issues that the CRLs have brought to light.

I'm thinking about Replimune's RP1, for example. This was, of course, a different administration and a different set of criteria used to judge the BLA, but at the time it was very clear that the FDA was, come hell or high water, not going to approve that drug under the old administration. They gave the reasons for that, and we were able to ascertain exactly what the FDA thought of that review at the time.

So, we may be entering a different era again.

Yeah. I mean, look, everybody wants to be able to see everything and read everything, but of course some of the companies have argued that this was one step too far. I agree that I would also like to see it, but this is what we do, right? We want to find as much information and detail as possible.

Eric, while you're on, do you want to also talk about Agios's priority review for, as you rightly said to me, a drug that missed its primary endpoint? What do you think is going on there?

Yeah, and I'd love to hear others' views on this new FDA and whether maybe we're swinging a little bit too far to the right in the other direction—the direction of lowering the bar and approving anything that looks active in rare diseases, despite maybe not having met the statutory requirements for proving efficacy.

What we have with Agios is a drug, mitapivat, for sickle-cell disease. These are obviously patients who are very, very, very poorly off and in need of new therapies. There are probably very few unmet needs that are any more substantial than this one.

Nonetheless, the drug missed its co-primary endpoint, a reduction in VOC pain crises. That data came out, I think, in November of last year. The stock got hammered, and everyone pretty much gave up on the program and the company. After talking to the FDA, Agios moved forward with the filing.

This week, we learned that the filing was accepted for priority review. So, not only was it accepted, but it was accepted under the auspices that there's a meaningful data set here and a meaningful unmet need. I'm told that an AdCom will not even be required here.

If anyone needed any sign or signal that the FDA is open for business with regard to pretty much any orphan or rare-disease opportunity, here it is yet again. I guess the question for this group is whether this is a good thing or not.

Obviously, it may be a good thing for sickle-cell patients who can try the drug and may get benefit from it, but that benefit has not been proven in clinical ways just yet. Of course, there are going to be costs to society from having another very expensive drug in the marketplace that hasn't fully justified its cost-benefit.

Yeah. Paul and Josh, you both cover sectors where there are companies with this type of difficult endpoint and rare-disease situation. Do you want to chime in here? Maybe Paul first, and then Josh.

Yeah, sure. Can you guys hear me?

Yep.

Okay, great. I want to throw something at Eric, too, at the end of this, because some of the situations that I'm following most closely here are, obviously, this uniQure saga, which was not a CRL walkback but seems to be a “you can't file now, you're more than welcome to file” dynamic, right? The stock chart follows that.

REGENXBIO and Hunter syndrome; what's happening in DMD. Sarepta got filings accepted for exon-skipping therapies on the back of trials that technically didn't work, right? Dyne is still filing on dystrophin.

I mean, we’ve debated this in a number of these. I at least feel like, to a few of the situations I’m close to, like the uniQure situation, I felt like they didn’t get their fair shot. I feel like that’s something that should get reviewed with an AdCom.

But I guess from my seat, I’m not yet assuming that all of this is just unequivocally going to go in favor of the sponsors that are getting back in front of the FDA. Maybe every situation is different, and I’d be curious what Eric thinks about Replimune at this point. Can it actually be rejected again?

But for uniQure, I still think we’ll have to see if there’s an AdCom. And I think the Street agrees, too, because the stock is around $40, which is probably pricing something halfway to what it would be if the drug got approved with a favorable label that people saw as de-risking a big commercial opportunity.

So, I think for some of these, at least getting their day to have it reviewed and assessed—especially the companies that invested money and ran trials based on a certain premise—is probably the only fair thing to do.

Do you think, Eric, in the situations that you’re following, like an Aiolos or Replimune, are those largely de-risked?

You make a great point, Paul. There’s a spectrum, I guess, of probabilities of success here that we’re talking about. Maybe the point that I was trying to make less eloquently was that that spectrum has shifted massively to the right.

If you used to assume that Replimune or Aiolos had a 10% or 20% probability of success, I don’t know. I’d probably put that well above 50% for both of those drugs right now, maybe close to 70% or 80%.

I don’t know where you are with uniQure or Dyne. I think that was always a pretty high-probability success. They would really have to walk back some of the regulations around dystrophin production to reject that drug.

For some of the others you mentioned, Regenxbio and others, my guess is they’re right-shifted by a meaningful amount, too. Do you guys think that less of a right-shifting is the right way of looking at this? Am I too optimistic in thinking these are 70% or 80% probabilities?

Let me just get Josh onto this, because I remember the discussion with Sarepta and whether it should have been approved or not approved, Peter Marks, et cetera. Josh, you had some strong views at the time. Do you want to chime in on this?

Well, to Paul’s point, it is a spectrum, and every program and product is different. I guess where I’m thinking about this, or the lens that I’m coming at it from, is wondering: We’re making decisions on behalf of patients. I think many kind of lose sight of that, but also lose sight of the challenge that represents.

Patients obviously want options. They want hope, and they’ll probably be willing to take more chances in pursuit of that hope. The FDA obviously needs to safeguard those patients from drugs that may offer more harm than benefit.

I think one could have argued around some of the Duchenne therapies that they actually did offer more harm than benefit, because the data package was really not compelling at all. But where along the spectrum do we shift from, “Okay, the FDA is going beyond protecting the interests of patients” to approving drugs that may not be detrimental but may offer some small benefit?

Everyone’s going to draw that line in the sand a little differently. That is probably what makes the FDA’s job so difficult, because everyone’s line in the sand is different. Whatever decision you make, there’s obviously the black and the white, but we’re talking about the gray.

Rare-disease data sets are often in the gray because you’re not able to run rigorous, placebo-controlled phase 3 trials. Due to the subjective component of it, there are always going to be disagreements about where the line should be set.

Anyone who’s tasked with that job is always going to be in a nearly impossible position. As a result, the more transparency that the FDA can bring to its decision-making process, the more it can at least try to align everyone around where that line in the sand is being drawn and why.

Matt, I’m going to come to you. I’m going to pass it on to you to talk about all the AI stuff, but open with this response, and then move on to telling us what you learned from Dario and everything else.

All right.

Oh, okay. Sure. The simple point here is just that I think a lax FDA is what we probably would have expected at the beginning of the Marty Makary era. I think that’s what you’d expect from the rhetoric from the administration early on.

I think what happened was full of surprises. But I’d also remind everyone that we’re dealing with an FDA that’s really been decimated and has lost a lot of seasoned people. That makes it a lot harder to, instead of being lax or tough, just get it right. I think that’s a very tough thing for people at the agency now who are facing a lot of these decisions, as a lot of people are gone.

Hey, Matt, can I ask you a question?

Sure.

Does the loss in experience at the agency make the agency intrinsically less flexible? I think about some of the rare-disease products.

Well, yeah. With uniQure, I think Nicole Verdun was a huge champion of that, is my sense. I don’t know that 100%, but it sort of makes sense. She was the deputy to Peter Marks, and it kind of fit into the sort of stuff that was said publicly there.

Then you wonder, when you go down a rung or 2 at the FDA, is there really going to be the champion who’s going to want to stick their neck out for something like this?

I’m less concerned about the power vacuum than just the fact that so many people are gone. I don’t remember the specifics—I remember the specifics for OCE—but I just know, when I’ve spoken to people who were recently at the FDA, they’re just like, “So many people are gone.” It’s just doing your job when you know the person down the hall that you went to for expertise, that you went to for counsel, is not there anymore, on a really big scale.

Starting with DOGE, before Marty ever came in, we were dealing with an FDA that had lost a lot of people. That’s really tough. Then you had this suspension of the normal process over the past few months. I just think they always have a really tough job, and all that makes it harder.

Certainly, if you had a Peter Marks or a Nicole Verdun, those were people who pushed for these applications, who had opinions and pushed for flexibility. I don’t know whether we land on, “Well, we’re in a right-to-try administration, and there’s a lot of flexibility,” or, “We need to fall back on the regulations.”

I think the administration’s biggest issue with the FDA is that they don’t want to be hearing about it in the news all the time. I think that’s how the FDA was traditionally managed, but now, after all of this mess we’ve had this year, I don’t know how it plays out.

All right, Matt, tell us about the AI.

AI is really cool, and nobody knows what to do with it. I had this amazing experience. I was asked to moderate a panel with Lotte Bjerre Knudsen, who was basically the leader of the development of Ozempic, and Dario Amodei, the CEO of Anthropic, when they launched this new product, Claude for Science, which is really built to help people in science, and in particular at drug companies, do work.

A lot of it is very research-oriented. This is really kind of an ultimate lab-notebook product that helps you think about your experiments, helps you draw your diagrams, and helps you analyze your data. It can keep in mind a corpus of all the stuff you’re working on.

There were also a bunch of big executives who were obvious AI believers to one level or another—people like Vas Narasimhan, Aviv Regev, and Chris Boerner from Bristol Myers Squibb—who were there talking about both why they’re big believers and what the issues are.

I do think there are 2 things we have to separate for the whole AI field, and for Anthropic in particular. One is: Is there useful stuff you can do now? I think the answer is yes, and people are going to learn more. There was one point where Chris was talking about 5% to 10% efficiencies for Bristol, but across the board. That sounds great.

Then there’s this big pie-in-the-sky stuff that Dario Amodei has always talked about, where this is going to change biology, where every year is going to be like a decade. That’s his actual prediction. He now says that won’t happen for at least a decade, and I think the jury is still out.

It’s certainly plausible that this helps industry pick better targets and helps lower the failure rate. That is the toughest job for AI, both when we’re talking about just LLMs, which are this specific set of tools, and when we’re talking about all of our favorite biotech AI companies.

I also got to talk to Mark Tessier-Lavigne a few months ago, and you know, what do you do with these technologies in order to make drug development better? We know that the track record of new technology in drug development is that it makes it more expensive.

So, Matt, these folks—Dario Amodei and Demis Hassabis—are luminaries in AI, and they do allow themselves to opine on a subject that is very complex, which is biology and the translation of biology. I do wonder whether—

I do wonder whether—at least Demis Hassabis from DeepMind has a Ph.D. in neuroscience, so he's got some background.

Well, I mean, Dario Amodei is in biophysics as well. The neuroscience led to the work, so they have some background.

But yeah, I think that's true. Do you think they—because this point often comes up that 1 year is going to be equal to 1 decade, and it's going to take us a decade to get to that point—do you think they really know what they're talking about? I'm not—I hope I'm not casting aspersions on anyone—but do they really know the topic well enough for us to be listening to them more?

Well, I think they know the topic well enough to have opinions that we can say are informed. I also think the opinion—I'm being a little bit of a journalist here—of somebody who's watched a whole bunch of technologies be adopted in biotech and had to see things get worse, or not get easier, is also a pretty valid viewpoint.

I was really impressed with Amodei. And I mean, Vas is on his board, right? Mhm. He does seem to have been listening to the thoughts about this, and he did recognize expertise. The point is really very much that he thinks this is going to be a multiplier in what a single biologist can do, which I think there is a good argument for.

The big question to me is whether the kind of reasoning these models do becomes the kind of reasoning that gives you a brilliant insight, how much of it is doing things in parallel, and how much of it is not. But Narasimhan made a comment as if it would be a minor thing that, well, you can't get rid of all of the time in drug development, but maybe you can decrease it by 30% and go from a decade to 7 years, and maybe you can go from an 8% success rate to a 16% success rate.

For any of us sitting in the room who've been looking at drug development for any length of time, those are astounding numbers. I mean, the point is you don't have to do what Dario says we're going to do for things to change a lot.

Yeah.

I think the biggest question in this field broadly—and not just about Claude for Life Sciences—is really just whether you can make your people in your research labs more efficient, right? Which is very much along the Claude Code path. They're really just trying to make another Claude Code; they had such success there.

Yeah. I mean, we call those apps, I suppose, in one way. I think a lot of people—or Google didn't make as much noise about this at the time—but a couple of years ago, maybe it was a year ago; I don't know, things move very fast in this world, they published a paper on Google's AI co-scientist, which is, I think, 3, 4, or 5 agents who hypothesize and check each other and then actually give you a biological hypothesis for a particular disease, which is quite interesting.

But I haven't looked either.

That is the kind of thing that Claude—you know, if you've used Claude for just writing reports, it is really great. You can go into this thing and say, “Give me a report on HPV in head and neck cancer,” because that's the last one I did. And if you actually know what you're talking about, it'll give you a report that's good enough. It's not perfect, but it's really fast and it's a lot of information at once.

Yeah.

And the idea is that you can have your 5 agents and they're arguing with each other and you can do all these things. My big worry for biology—and I did ask them about this, and the answer was, “Yeah, that's the problem”—is that with coding, you get an answer pretty fast about whether or not your code runs right. There are some early tests to see if it works.

The whole problem with drug development is you can be a brilliant person who can spend a decade chasing the wrong thing, spending hundreds of millions of dollars, and it turns out it was the wrong approach the whole time. You don't really get the same level of checkpoints along the way, certainly not in the way you do if you're running code.

And I think that is a big issue. But then you have to look: there are a lot of things that are very rote, like regulatory documents. So are clinical-trial enrollments. So are keeping track of all the stuff in your lab notebook, making sure everything is in order, and coming up with new hypotheses. Could these tools be additive? They could.

I mean, my worry, if I were managing a lot of people with them, would be: how do you differentiate between people becoming more efficient using AI and people doing what I think we all do when we first encounter the tools, which is you sit and argue with Grok for a few hours and never get that part of your life back? So I think there's a management challenge.

Yeah. Yeah. Absolutely.

Neat.

It is. It is, and we're all waiting to see the productivity gains that the pharma companies talk about. I had a podcast recorded with Christian Machesi [?] from Bristol Myers Squibb where he talked about the same numbers: a 30% reduction in the time for drug development. We need to see some of those things come through.

But, in the interest of time, thank you, Matt, for that. I'm sure there's a ton more that we could talk about on this. Let's talk about some follow-ons and M&A, et cetera. Paul and Josh, who wants to pick that first? One of you go. Yeah. Okay.

I'll go. Okay. Yeah. Yeah. Yeah. So, I cover Vertex, and maybe Josh covers Crinetics. I can at least give a little bit of an angle on the Vertex side.

So, Vertex this week bought Crinetics for close to $10 billion, closer to $9 billion on a net-cash basis. Crinetics is an endocrine company, and this sort of specialty rare endocrine space has quietly become pretty hot. It feels like it's the next little niche space to get hot and become its own little subvertical, kind of how the renal space has over the past 3 to 4 years.

Crinetics has 2 key drugs, paltusotine and also atumelnant, and a lot of the discussion on the call was about the second drug, which is in Phase 3 development for CAH, congenital adrenal hyperplasia. Neurocrine has validated that that is a huge market via their drug Crenessity, which—you know, I cover Neurocrine; I've covered it for a long time. I think Josh has covered it for about as long too.

I feel like it's one of those drugs that works decently well, but really the big selling point is that it's super safe and easy to use. This disease is a big unmet need. Patients have been using steroids for 50 years to treat it, and it's on newborn screening, right? So, if you're a rare-disease company, there are a lot of patients out there right away that you don't have to find.

The only real debate I've heard from the Vertex side in this deal is just: did they overpay? It was almost a 100% premium. Crinetics is a stock that had kind of lagged the tape, in part because of long timelines and things like that.

I almost don't know how much it matters to Vertex if they overpaid by a little, as long as there isn't a big issue with atumelnant. I think the only sort of risk question there is on the liver-safety side. Vertex talked about 7 liver-enzyme elevations in the program so far, all mild and self-resolving. That was a higher number than I had heard before, but you kind of have to give Vertex the benefit of the doubt.

If there's anything that Vertex has extremely outstanding expertise in, it's small-molecule drug development. Assuming they got that point right, it feels like the floor value for both of these drugs is probably $1 billion, if not more. Vertex is saying $5 billion combined. I don't really know if you have to believe that, but I think, broadly, Vertex is at this point where they have the kidney angle beyond CF, and this adds another important angle to a company with a $130 billion market cap. This is the kind of stuff they should be doing, even if people debate the price of it. What do you think, Josh?

Sad to see Crinetics go. Endocrine is one of my favorite spaces because there's essentially no target risk. There's very little in the way of competitive dynamics, and there are large unmet medical needs.

Crinetics was really just at the start of what I thought was going to be a very, very sizable effort across its programs, including its preclinical programs. When there's no target risk, you can actually start to look into those preclinical programs and fairly easily envision a clear path through development and regulatory review and ultimately commercialization.

The only real risk you take, to some degree, is medicinal-chemistry risk, and Crinetics had proved itself quite adept. So, kudos to Scott and his team—a wonderful group in San Diego. Within a year, we've lost Vividion and Crinetics from the San Diego ecosystem, which hurts, but there's still tremendous innovation happening here and across the world. So, personally, I think everyone benefited from the deal.

I think it's time to move on, Josh, and see if the deals stop.

Yeah, it’s come to London. There have been a couple of other deals that are quite interesting. Over the past 6 to 9 months, there’s been extra activity from European buyers, including some private European pharmaceutical companies and some public ones. The latest round was Ipsen buying 2 companies and paying decent amounts of money: an upfront payment of $450 million for Kartos Therapeutics, which is a U.S. company, and then a €200 million cash deal for Memo Therapeutics.

Both companies have assets in phase 2 or phase 3. One of them, Memo, has an asset in renal disease, which is quite interesting for patients with renal-transplant complications, and the first one, Kartos, is in myelofibrosis, which is in phase 3 with data expected in 2027. That continues a spate of activity by UCB and Recordati. If I’m not careful, I’ll probably forget the whole list.

I didn’t make a point of listing all the companies. Some Europeans are taking advantage of this M&A window—or feeling a bit more aggressive with regard to their ability to obviously finance these and gain access to them.

Another interesting thing was Novartis buying Myricx, which is an ADC company. It has an interesting story. I would go and read about how this company, which was seeded by Sofinnova Partners and Brandon Capital quite a few years ago in the U.K., started off developing drugs for oncology and then changed its strategy and moved into the ADC space. Novartis is paying $1.1 billion in cash upfront.

It’s a private company and it’s not in the clinic yet. There’s a total of another $400 million that can be paid. It felt a little bit rich for a private company, so there must be something really exciting in there for Novartis to have done this. This is the second ADC deal out of Europe.

The other one, of course, as everybody knows, was Gilead buying Tubulis, which was at a different stage of development, and Gilead had already had a deal with them for a while. So they probably had quite a lot of knowledge with regard to the drugs being developed, or the linkers, et cetera, that they had. We keep being drawn to China, looking at deals that people do for ADCs, and of course, here are 2 that happened based out of Europe.

To Josh’s point, the capital markets in Europe are very tough. I don’t know if either of those companies would ever have considered listing in Europe; they would probably have gone directly to the U.S., but now they’re owned by others. Hopefully, we’ll see some of the results of these drugs that they’ve been developing come to fruition.

In terms of other deals, we had a raise this week—or last week; remember, we’re covering 2 weeks here—which was BridgeBio’s $1 billion financing from Sixth Street and KKR, a health-care royalty deal that obviously provides capital for the company to go ahead. Josh, you look at BridgeBio, and of course this happened, and then earlier this week we had a failure of the clinical trial, which we’ll talk about later, of AstraZeneca’s cardiomyopathy drug. Do you want to touch on that, and do you think they regret waiting?

Well, no, the eplontersen updates are far more relevant. BridgeBio is about to launch 3 new drugs over the next 12 months, and that’s obviously going to be capital-intensive. The launch of their product Attruby, which we’ll talk about as we get to the ATTRibute data, has been exceptional. But they’re still burning cash, so it’s prudent to provide just a little bit more of a buffer as you go into 3 new meaningful product launches.

Yeah. It didn’t look like a particularly expensive deal in terms of the numbers in there. So I think this is a perfect move into the data conversation that we’re going to have.

Let me start off with the AstraZeneca news that came—as I said, the weeks get all squished into 1 day in my head. Sometimes I think it was yesterday. AstraZeneca’s eplontersen failed to meet the primary endpoint in a late-stage trial in ATTR cardiomyopathy, and the stock was down 10% yesterday. There was a lot of conversation in our drug chat and so on about whether that’s fair or unfair. Bloomberg TV—I might have used the word “unfair” a couple of times.

It does have an impact on a whole bunch of companies that the good folks on this call cover. When you have a major trial readout like this go in the opposite direction from where I think everybody was expecting—and Astra was quite positive about it, too—it’s not a surprise to see such a big share-price move. On the other hand, it doesn’t impact their $80 billion 2030 target. Of course, it takes 1 option out of the equation to a degree.

How this gets developed going forward in a subgroup will depend on what that subgroup data is. I’m sure Josh, Paul, and Eric will comment on this. Part of the problem for Astra now is that this was supposed to be the easy one, and now we’ve got AVANZAR, which is TROP2 in lung cancer, and SERENA-4, which is an oral SERD in breast cancer. Both of those have people worried, and they’re coming up during the second half of this year.

I think that’s why some people may be reticent to be the buyer in this equation of buying and selling the stock, which is what we’ve ended up with. Potentially, the pressure on the stock wasn’t there to be absorbed by anybody else. You’re looking to a second half where there is risk with these trials, which we’ve written about very openly, very much on the Terminal. So that is perhaps 1 of the reasons why the stock has been under so much pressure.

Nevertheless, I think some people in the investment community view this as a buying opportunity. Some analysts have even said that. So let me pass it on to Josh, to start with BridgeBio, and then we’ll talk about Ionis. Of course, we know what happened there. Josh, Paul, and then Eric, jump in as you want—and Matt, of course.

Yeah, this was surprising and a head-scratcher. I’m looking forward to getting the full data presentation next month at ESC to figure out exactly what went wrong. One of the obvious explanations is that nearly 60% of patients were on tafamidis at baseline, and another approximately 25% went on a stabilizer during the course of the trial, either tafamidis or acoramidis.

Even with that, at least in theory, adding TTR knockdown in addition to the incomplete stabilization from tafamidis should have had some type of separation. The press release suggests that there was no effect. We’ll have to see what that actually means. Does that mean it was not statistically significant? Does it mean there was a trend, or was there literally nothing going on?

If that’s the case, now we’re going to have to figure out what that likely reflects. Could it be that tafamidis is a better drug than many of us thought? That seems very unlikely, just given the totality of the data. Could something else have been going wrong with eplontersen? We’ve known that ASOs can have some cardiotoxic element, so is it possible that manifested in this trial for some reason? There are lots and lots of questions.

For the time being, though, it looks like it’s going to remain a 3-way market battle between BridgeBio, Alnylam, and Pfizer, with BridgeBio and Alnylam really picking up the most market share with more differentiated product profiles. Pfizer is also hanging in there with its massive commercial sales force and ability to find patients, maybe before they get into the fold where they could consider Attruby or Amvuttra, and start them on tafamidis first.

So there are tons and tons of fascinating moving parts for this unmet medical need. It’s great to see that there are terrific options for patients and that the bar now does seem to be raised so high for care that it’s hard to improve upon it. Again, we’ll have to see the totality of the data to anticipate what this might mean for other TTR cardiomyopathy programs.

There are some who think that this means we may never see another positive TTR cardiomyopathy readout again, because outcomes are just so good. That’s to be determined, though.

Josh, can I chime in with a couple of things on what you just said?

Yeah.

Yeah, thanks. I caught up with the Ionis team yesterday, and they confirmed that in the tafamidis combination group—which doesn’t include the drop-ins, but just the patients on tafamidis at baseline—“no benefit” means no benefit. It doesn’t mean that there was, say, a hazard ratio of 0.88 with a p-value of 0.4; it sounds like that hazard ratio could be pretty close to 1. They also said there was no benefit in the secondary endpoints.

Here’s the strangest thing: there was no benefit on cardiovascular biomarkers. There wasn’t even a biological effect. I think what’s really confusing about that gets into the implications going forward. Do physicians look at these data and influence treatment decisions? Does this impact the likelihood of success for Alnylam’s really important trial for nucresiran?

Alnylam’s HELIOS-B study, which read out a couple of years ago, showed a really substantial outcomes benefit on top of tafamidis with a silencer—maybe a slightly more potent silencer, but not more potent enough to suggest a difference of 0.2 in a hazard ratio. But it also showed a pretty significant biological effect on things like NT-proBNP.

That, to me, is really confusing. We had spent a lot of time thinking about the powering of the tafamidis combination group in this study, because that was 1 way the data could have differentiated from others. I think we were concerned that even if there was an effect, it was somewhat underpowered, because as patients are treated with tafamidis earlier and earlier with this disease, the outcomes improve, as with pretty much any condition.

There have also been these noninvasive scans where patients are identified earlier, and I think that’s probably a piece of this. But it doesn’t feel like it can be all of it. I still think, going into ESC later in August, it’s very confusing that this sounds like it was really a dud on top of tafamidis.

And to Josh’s point on investors wondering whether there will be another positive trial in TTR cardiomyopathy, Alnylam’s stock yesterday opened up about 17%, based on the idea that, okay, now they’re going to own this silencer market. The interesting nuance with Alnylam’s valuation and the longevity of this franchise is that Alnylam owes a 20% to 30% royalty to Sanofi on its drug Amvuttra, which runs a 50% to 60% operating margin. That royalty is a huge element of the NPV and the economics.

A key thing for Alnylam is eventually having its next-generation Nucresiran, which doesn’t have a royalty burden, succeed. It has a convenience advantage because it’s dosed every 6 months. But Nucresiran is in a cardiovascular outcomes study right now in ATTR-CM, and Alnylam hasn’t given numbers for it. The perception is that the vast majority of patients are going to be on tafamidis at baseline, so it’s essentially an adjunctive study. The question is, does this Ionis readout with AstraZeneca undermine the TTR-silencer synergy in a contemporary population, or can Alnylam argue that they have that on their label?

My thought, assuming my read on what was said yesterday is right, is that there really isn’t even much of a trend on top of tafamidis. You have to wonder if the most logical thing for Alnylam to do will be to find some way to expand the sample size of the study and try to enroll in territories where they can get more monotherapy patients. At face value, the monotherapy data for eplontersen, with a hazard ratio of 0.71, looks very, very similar to Amvuttra. So it’s very confusing here. I think a lot of investors thought this study would work, but maybe they wouldn’t hit a P value on top of tafamidis. To show nothing is—I don’t know. I think it’s pretty shocking.

I’m still a little bit dubious, though, as to whether this data will, outside of removing a competitor, influence the prescribing of Amvuttra or Onpattro. I don’t know if you had a view there, Josh. That’s one thing people have been asking.

Yeah, we spoke to a couple of doctors yesterday after the data. They weren’t going to change their practice, but they also weren’t the ones who were prescribing combination therapy to begin with, which makes unbelievably little sense for the most part. Once you have either Amvuttra or Onpattro on board, you’ve nearly fully eliminated the monomers, and so there shouldn’t really be any incremental benefit to be had.

It’s a little bit different for tafamidis if it is indeed a less effective stabilizer, in which case the addition of a treatment option should have added benefit, which is why this whole data set is so confusing. At least from the checks we’ve done, there was no major difference from those doctors. But they absolutely thought that payers were going to start pushing back much harder than they already have around the combination use.

On the combo?

That said, it may be difficult because, in some payer dynamics, the Part B and Part D payers are operating independently. The left arm may not know what the right arm is doing in allowing 2 drugs to be used at once.

Right. And I think you know this really well, but I think that’s a big reason why tafamidis has been able to sustain pricing at a really significant premium. This whole cross-management dynamic, which seems so obvious to an analyst or investor who’s deep in the weeds and specifically in the TTR space, just isn’t really done at all major payers.

I was just going to close out and say that we’ve talked in the past about some of the very sloppy thinking amongst cardiologists when it comes to treating ATTR cardiomyopathy. What’s going to be really interesting is, as claims-data analyses are done over time to provide more evidence for the decision-making process and tighten it up, it could have some really profound implications. I’m keen to see what happens there.

Just in the interest of time, let me highlight that John Maraganore did say something on Twitter about RNAi versus antisense. I don’t know enough about it to know why one would be better or different from the other, but it relates to what you were saying.

Yeah. Can I comment on that, Sam?

Exactly.

I actually emailed with John yesterday, too. The Alnylam drug Amvuttra and, even more so, Nucresiran get better lowering of the TTR protein, with less variance, and attain their peak faster. Those are definitely true, and I think John makes a fair point on that.

The reality is that the Alnylam drug is a little more potent. I think siRNA for almost every target seems to be more potent, outside of some of the things that ASOs have accomplished, such as in the CNS. That’s fair. But again, we’re talking about 1 study that had a 33% reduction in mortality for the silencer on top of tafamidis, which is mind-blowing—almost too good to be true. But the data is the data.

An older Alnylam study, the APOLLO-B study, showed a trend on top of tafamidis on outcomes. Then this study with an antisense drug that maybe lowers the protein 5% less, with greater variance and slower onset—but how much does slow onset matter? I mean, it’s a 33-month study, right? It doesn’t seem to show any benefit.

I think John makes a good point that the siRNA drugs in this space are more potent, but if I’m Alnylam, I can’t bank on that and I can’t bet Nucresiran on that. The Alnylam team is obviously super sophisticated, and I would imagine that, after seeing the ESC data, if it looks how a lot of people suspect, they may try to do some things to mitigate the risk in this Nucresiran study.

All right. Thank you, guys. This was a nice deep dive on this. It’s a very, very important drug and very, very important for such important companies.

Let’s move on to data that also came out from Revolution Medicines, which seems to be having a nice run of positive data. They reported Phase 1/2 trial data for zoldonrasib, which is the RAS G12D-selective inhibitor. Remember, daraxonrasib is the pan-RAS inhibitor that we’ve seen so much excitement about.

This was in 2 combination chemotherapy regimens for first-line metastatic pancreatic ductal adenocarcinoma: an 82% overall response rate and a 96% disease control rate. These are fantastic numbers, right? That was with FOLFIRINOX. Then, with gemcitabine plus nab-paclitaxel, the combination was a 61% overall response rate.

In general, the safety profile was manageable. Of course, a lot of oncology trials say that, but when some of us see the results, we go, “Oh my God, how is that manageable?” These results again confirm the potential value of this in the first line, and it’s advancing to Phase 3. So, another positive for Revolution Medicines. Eric, I don’t know if you wanted to add anything to that.

Not really, Sam. I think you covered it well. I do think that zoldonrasib looks good. I guess the question now is, in a daraxonrasib world, what incremental benefit does it provide? Daraxonrasib is so good.

Yes, zoldonrasib, for those patients with G12D mutations, is going to be a little bit more tolerable. But what caught my attention was the combination of zoldonrasib plus daraxonrasib—a G12D inhibitor plus a pan-RAS inhibitor. It didn’t look so much better than pan-RAS alone.

I’m not sure there’s synergy between these 2 drugs, and that may open up the opportunity for other drug combinations—drugs that have orthogonal mechanisms, not RAS plus RAS, but RAS plus non-RAS—to work better. Of course, the Tango data that we’ve discussed before is probably much superior.

Yeah, that’s going to be quite interesting to see. But I also wonder whether some physicians and some patients may just not want to go on daraxonrasib because of some of the side effects. At least if they have G12D, it gives them an option to not necessarily have to take daraxonrasib if they’re scared of the side-effect profile. I don’t know. This is a very difficult disease, so it’s going to be quite interesting to see how it pans out.

The next data we had, I’m going to go back to Paul for, is Skyhawk’s Huntington’s disease data. After that, we’ll go on to AAIC, which is coming up next week in London, my hometown. So, Paul, do you want to just quickly touch on this Huntington’s disease data, or would you rather move on to AAIC?

Let’s do AAIC, actually, if that’s okay.

Go for it. You and Eric are going to own this.

Yeah, I’ll start, and I’ll be brief because we can hear more from Eric. I think I saw a headline that maybe Eric is a little more bearish here, so I’m psyched for some spiciness.

Basically, the big thing at AAIC coming on Tuesday is that Biogen is going to show full data for its anti-tau ASO for the treatment of Alzheimer’s. The study missed its primary endpoint, but Biogen is moving it to Phase 3.

In my number of conversations with Biogen since this data, they are trying to convey enough enthusiasm about this so that they are defending and standing by their view that it makes sense to move this to Phase 3, while also trying to manage expectations into the data.

The oddity here is that the primary outcome was a dose-response analysis. So it required there to be a linear dose response, or some degree of linearity to it, when it sounds like the lowest dose did the best. At face value, for a drug that is working at the gene level and trying to reduce the production of tau so the brain can clear it, that doesn’t really make sense.

I think there are reasons to think that there could actually be a true inverse dose response, or maybe a nonlinear one. One possibility is that as you push the dose, maybe you get some sort of idiosyncratic safety stuff with the oligo. We've seen some oligos have dose-limiting toxicities when they're pushed in the brain, right? Tofersen's one; there was an ALS one. Who knows? We've seen this across other companies, too.

Maybe the actual tau knockdown just has a wider degree of variability across doses because this is intrathecal, and the biodistribution is going to be highly variable from patient to patient. Biologically, I love this modality and approach, and that's why I think I'm a little bit more hopeful that the data look interesting. That's really because I think tau has the strongest scientific rationale for a disease-modifying target in Alzheimer's. It correlates with progression. It also precedes progression in different areas of the brain; you can stage the disease based on tau aggregation. And we know the antibodies don't really work because they don't actually lower intracellular tau.

The question here is really: Is the data good enough to get people excited about the Biogen approach, which is intrathecal and presents a whole other challenge? Or is this data basically showing some level of Biogen's tau proof of concept, but then the more interesting plays are the brain-shuttle approaches, like Denali and Arrowhead, which can be given IV and just have a different kind of risk-benefit profile, and maybe a much lower efficacy bar given the ease of use? But, yeah, Eric, what's your view and prediction on the degree to which this data is encouraging or super messy?

Great summary. Thanks, Paul. I kind of agree that tau is the best—maybe the best of the worst, but the best target we have. That doesn't mean it's a good target, or that it's a good use of capital for Biogen to move forward. But I guess I'm curious what your feedback has been.

You're right, we've been kind of critical of this decision, and maybe folks are preaching to the choir when they call me, but in terms of my buy-side interactions, it's like 9 to 1 sort of negatively inclined toward Biogen's decision to move this forward. When I speak to people, they just want this to be over and move past it, and the sideshow that Alzheimer's has been for this company for so many years now to maybe hopefully go quiet or quiescent for a few years as this Phase 3 program begins. Is that the feedback you're getting as well?

For the most part, I think there are some people who like this and maybe don't want to admit it to a sell-side analyst like me because they don't want to talk theoretical credit into the stock. But, yeah, I certainly think the Biogen bull case has become this broader pipeline optionality thesis, and you've just got lower-risk, more proximal opportunities with the 2 lupus drugs and felzartamab and things like that.

I just thought, Eric, when I talked to Biogen in the past year, they were pretty cautious on the commercial setup for an intrathecal therapy. Because of that, it made me think that if they were going to advance it, the data just can't be a total mess. But we'll see—the beauty is in the eye of the beholder, right?

Yeah. I also don't get the sense from Biogen, even this week, that they want to break the bank on this opportunity, right? They're kind of playing it middle of the road. They're using words like, “We'll see what the data show. It is what it is. We know we're not going to convince everyone.”

So it'll be curious to see what they are seeing, because you're right: They're making a massive bet in terms of expense, time, and effort to go through a Phase 3 program with 2 very large randomized studies. I think there'll be something, but I don't know. It's hard. I guess you and I both have PTSD from Alzheimer's now, and it's just hard to have much faith in any Phase 2 program.

It has been a rough ride with Alzheimer's, and tau is, I would say, a favorite. It'd be nice to see some good data, but I hear all your arguments and I agree with both of you, which is quite fun. Yeah, go ahead. You can have 2 sentences.

Well, I just want to say that in a world of uncertainty, where the whole world has seemed to be much more difficult, in many ways it is just really nice to be hearing about an Alzheimer's meeting where we're arguing about trying to read positive results from a Phase 2 trial and hoping we can go on to Phase 3. It makes me feel like there is some continuity in the world. So I wanted to thank you guys for that.

Very funny.

That is good. Well done, Matt. That was a really good way of ending the call today. There's a whole host of topics. Everybody knows my list was enormous, so I'm going to leave it at that.

I want to remind everybody to look at the Avidity story. What a lovely turnaround. You might agree or disagree, or there might still be a risk, but the company got a data set that worried everybody at the beginning of June, and they said, “But don't worry, we'll show you the details.” They came and showed us the detail. The share price completely corrected. I'm talking about a 30%–40% share-price move here, right? And then, of course, they've gone and raised $920 million on the back of it.

So congratulations to them, and hopefully, as this data gets presented, we'll get to see much more detail and see how it all works out. That looks like we're going to have another successful European biotech here—unless it gets taken out, Josh—which is good to see. I'm going to end on that positive note.

Thank you, everybody. Thank you to my co-hosts, Josh Schimmer, Eric Schmidt, and Paul Matteis, and special guest Matt Herper from STAT. You can find all our old episodes on biotech hangout.com.