The FDA, Unicure, and the Limits of Accelerated Approval

Episode Summary

Anish sits down with Adu, a med student and biotech investor, to work through the FDA's contested handling of Unicure's AMT-130 — a gene therapy for Huntington's disease delivered via stereotactic brain injection. They debate whether the underlying data justifies approval, why the agency's mid-course reversal has rattled the investor community, and what the Sarepta precedent should have taught everyone involved. The conversation broadens into a bigger question: given that desperate patient populations will always demand access to anything showing a signal, who is actually best positioned to make the call on whether a drug works — the FDA, the clinician, or the market?

Chapter Markers

00:00 FDA approval of AMT-130 and investor reaction

01:16 Unmet need and the case for regulatory flexibility

02:37 Sarepta, Duchenne's, and the cost of approving under pressure

05:09 Accelerated approval done right: the Amylyx example

09:14 Debating the AMT-130 data and the historical control problem

13:53 Why stock price matters for trial funding

17:20 How Prasad could have changed FDA culture differently

19:37 The FDA's role from Kefauver-Harris to today

22:26 Competing Huntington's therapies in the pipeline

25:39 Prasad's tenure: what worked, what didn't

28:27 Media coverage of the FDA and science journalism

Co-Host Handles

@anish_koka and @drdigiorgio

Show Handle

@drsloungepod

🔗 Connect with the Hosts:

• Dutch Rojas on X

• Dr. Anthony DiGiorgio on X

• Dr. Anish Koka on X

• Dr. Dan Choi on X

• Dr. Sanat Dixit on X

Transcript

Anish Koka: All right, Adu Subramanian thank you so much for coming on. Adu to my attention as some people, as many people do in ⁓ line of what I do online. ⁓ Adu had very interesting takes on hunting disease when hunting disease was in the news a lot because of Unicure, which is a company that had a therapeutic that the FDA was maybe, maybe not approving. I was saying that it's a completely decision to me for the FDA to make. And it sounds the investor community is also, you know, was also surprised. But ⁓ once that happened, that kind of ⁓ essentially ⁓ FDA breeds life into this company because now it's OK, ⁓ you know, this is the valuation it's based on. ⁓ were your thoughts? And that created quite a firestorm of controversy. so through all that thicket of noise, ⁓ do ⁓ a couple of very interesting pieces of insight that clearly he has somewhat of a deep level of understanding. So I reached out to him. So it turns out I do, are a medical student. You're finishing your third year. Congrats. And ⁓ so what medical school are you in? that when that happened

Adu: You can actually go back on, it's really interesting if you map out the FDA events to the actual stock price. Like they report these two year results in July. The stock is at 350 and then it doubles to seven because these results, they're not terrible. They're not great by any means, but it's better than the placebo control data that they've previously presented. And then at seven, goes to like call it 17 after the

Anish Koka: Yeah. Yeah.

Adu: Yeah. I'm at Georgetown right now.

Anish Koka: Georgetown, okay, excellent. ⁓ so you're in DC, Wow, you're right next to the FDA, my goodness, right next to the heart of all the excitement. you went to undergrad, correct? ⁓ did you go to undergrad?

Adu: FDA alignment and kind of stays there until the final results. I think that people weren't expecting the FDA to align with this because other trials require like, require a placebo controlled cohort, typically, and even drugs looking for accelerated approval or having a little tough time like with PTC therapeutics. ⁓ think most investors, obviously, I think most people would be, would have been surprised Yes, went to UC Berkeley, went to UC Berkeley, studied biomedical engineering over there, and then came to med school directly from after like straight

Anish Koka: Yeah. I see. Yeah. Yeah. I see. So no gap year or anything like that. But the final year at Berkeley, were doing, you were doing something interesting that, leads you to us talking

Adu: based on the data set and the comparable companies. But the trend of the FDA in the last, how many ever years has been towards more flexibility for unmet needs. part of the reason they probably did this is that if you asked a patient whether or not they'd get this drug or they'd take this drug because of the signal, there's 99 % chance they'd say yes, right? Like if you describe the data, I'm sure there's a huge latent demand as we've seen after they've pulled it. Leads you to biotech. Yeah, so when I came into Berkeley I was Investing a little bit and then started investing into more into medtech type stuff because of my engineering background and then ultimately the end of ⁓ career at Berkeley that last six months Started to work a little bit more intensely with the hedge fund like investing in biotech specifically and then just continued to write and analyze stocks and companies and There seems to be a huge latent demand for any type of drug that shows any type of signal. Hence, the FDA flexibility can somewhat be justified, even if you didn't think the drug works perfectly, and this was just a signal-finding study. things like that because I'm a little bit addicted to it.

Anish Koka: All right, so tell me, just to give a little bit of a background is that there's a disease, pretty terrible called Huntington's disease. It's a neurodegenerative disease, autosomal dominant. You're fine until your 30s, maybe 40s, even maybe 50s. then depending on what phenotype of disease you have, you get... So you think a patient community that's desperate, that's dying, if they that, give me this drug because there's a signal, then that's a reason for the FDA approve it.

Adu: I think it's certainly depending on the severity of the disease, you could make the case, right? I think, you you have certain pediatric deadly diseases in which you're approving drugs based on single arm, five patient trials. You're looking at the plausible mechanism pathway in which you're potentially approving drugs based on a few patients and a plausible mechanism because the unmet need is that high. I think that

Anish Koka: progressively worse and everyone dies. It's a neurodegenerative disease. It's pretty darn terrible. ⁓ Of course, usually by the time you've found out have the disease, you've had children and ⁓ it's passed down autosomal dominant. And so you only need one parent to give you a gene for you to ⁓ have this disease.

Adu: It is part of the decision making to go into it. And it's just a matter of perspective on how much flexibility you want to go, how much flexibility you want to give.

Anish Koka: Really quite a terrible disease. the controversy recently was about a company called Unicure. Tell me a little bit about Unicure since you're ⁓ the here. Right, right. I mean, I think it's problematic because of course a desperate patient population is going want anything, right? So a desperate patient population during COVID, you know, consumed all sorts of different things that were, you know, didn't work. Some of the things that were harmful. And we've seen over and over again over the time course of history of medicine is that, oh, by the way, I mean, we don't have to go back that far, right? Look at Sarepta. Sarepta, which is Duchenne's. Here you have, you know,

Adu: Sure. Unicure was developing this gene therapy. was a locally delivered gene therapy using an injection and a brain surgery into a specific part of the brain. And the whole goal there is to silence a exon, a specific gene basically, to slow the progression of Huntington's disease. think it's been a long saga for them.

Anish Koka: Again, Sarepta is a company that has some, don't follow, again, I don't follow the space. You know, I follow the space because I was interested in some of the regulatory decisions that were happening. And I was actually started it because I was very interested by how the biotech community, the investor community and the science journalists covering it. So that made me ⁓ attempt to to verify what ⁓ was happening. so I looked into Sarepta. I looked into, you know, their exon skipping therapy. looked into, ⁓ you know, the

Adu: You know, you'd go back all the way to 2021 and they started this trial and the design of that trial has gone back and forth, ultimately culminating with now three year results. And I would say a fair bit of controversy from the FDA and patient communities on whether the drug works and whether it should be approved and how it shouldn't be approved.

Anish Koka: caliph kind of decision. ⁓ kind of dovetailed into this decision on ⁓ the gene therapy. And here you have a clear case of children that were dying. And again, ⁓ Marks made decision, ⁓ even though the that was done that wasn't that good to begin with, really didn't show benefit. it have a, does it have just one drug? Is there only one drug that they have?

Adu: No, they have a couple other drugs, but I would say that the primary value driver is this drug, this AMT-130 for Huntington's disease.

Anish Koka: Okay, and what does it do? do, yeah.

Adu: It's a, it's essentially, it's a gene therapy. So it silences the production of mutant Huntington proteins by specifically targeting exon one a, which is like the path that in theory, the most pathogenic part of the mutant, mutant Huntington, MRA. and the way it's administered is like basically injection directly into the brain under MRI guidance.

Anish Koka: in the patients that were in the trial, didn't show that. Then he, because there's nothing else and the community is demanding it, it ⁓ non-amuletary kids, right, who were older and then died. mean, so look, and that, so just from pure, ⁓ you know, trying protect patients ⁓ when they're desperate will try anything, that's a problem. Yeah. What are you, do you have a, terms of original approval ⁓ or the original trial, do you want to tell me about the original trial in terms of what they've done?

Adu: Yeah, so I think it's good to look at the landscape of Huntington's disease. Like if you zoom out a little bit, all the other trials, if you zoom out like one more step, like, it's a huge unmet need, like you said, because what ends up happening is that you see oftentimes your parent go through something like this where they progressively decline from age 40 or 50. And then it's a pretty horrifying disease to watch at the end.

Anish Koka: If you talk to the patients, even after for the family members of these of these young boys, again, Duchenne's another horrendous disease that they they they to they feel like, look, we did we did everything we can. You know, like there was a nice there was a stat news interview of Duchenne's mom as like, know, well, yes, kid got this and ⁓ LFTs were crazy he was in the hospital a bunch and he was getting steroids and now he may be getting on immunosuppressants to try to ⁓ abate the immune from. Yeah, sure.

Adu: And then you know that it's coming for you as well because you've seen that happen. But every other trial in the space is running like a randomized control trial. They have a sham group, like a placebo group, and then they have a treatment group.

Anish Koka: you know, the of the vector of Elvira's. But she's totally OK it and hey, look, maybe there's So, you know, the issue is you really to I mean, the question is, is the society what what should like how should we approach And we should that extremely vulnerable people are going to, you know, want things they may not be the most in terms attempting to decide whether or not, you know, stuff

Adu: Unicure wanted to imitate that basically. Like they had a randomized control trial as recommended by the FDA back in 2021. And the whole goal was to like try to accelerate the process a little bit by understanding what was the effect relative to placebo, both like functional benefits and also the lowering on the target biomarkers like neurofilament light chain, which is like a marker of brain damage and mutant huntington protein, which is in theory the pathogenic marker.

Anish Koka: works or not.

Adu: Well, I think that's why we have the accelerated approval pathway for these progressive diseases. Because let's say that you have a confirmatory trial for Unicure. This is where I think that it gets a little bit murky specifically for Unicure is that if you have confirmatory trial for Unicure and is ongoing and you approve this on accelerated approval and the trial is going to read out in a year, maybe two, and you say, hey, if this trial succeeds, we'll keep it on the market. If it doesn't. And that was back in 2021. And they specifically talk about this where it's, know, the FDA wants them to run this trial. They want to run this placebo control trial and have this like one year cohort specifically so that you can accelerate the registration process. Like this current trial that we're talking about was not meant to be the one that the FDA would approve. The drug on was meant to be more of a signal finding trial originally. That's where it started. it's two years and ultimately it's going to be a safe enough therapy and it's up to the decision of the doctor and the patient rather than the regulator on what is safe and effective for you, which I'm okay with. I think that I'm okay with placing the onus of finding what is safe and effective on a neurologist and their patient rather than, you know, one person at the FDA or a group of people at the FDA. If you're willing to pull the drug off in two years, I think Amalix is probably the best example of this.

Anish Koka: Right, wait, so the original trial was a sham control trial, right? And so what are the results of that?

Adu: Yes. So that didn't work. I mean, pretty straightforward. If you look at the functional benefit over one year, was no difference between placebo and drug. If just looked at that, you would say, ⁓ this...

Anish Koka: But then, at, okay, I know Amelix, but look at what happened at Sarepta. I mean, ⁓ attempted to pull Prasad and the Prasad FDA. don't know if Prasad, I'm sure guided that, but the Prasad FDA attempted to pull Sarepta off when you had ⁓ debts that happened. And then look what happened. Look what the investor community did. Look what, you know, Stat News and company did. I mean, they still tried to roast him ⁓ and ⁓ his firing. What was the end point they were looking at?

Adu: It's the composite Huntington's disease rating scales like C-U-H-D-R-S. It's basically just a four... for a category scale that measures things like reading time, movement, functional capacity. And it's a functional scale. So there's obviously noise within that because, you know, one day you might wake up and be able to move a little bit better than three months later because it is not a, you know, immediately deadly disease. You might feel a little bit better because you've parked a little closer to a car, right? We're parked a little closer to the office when you walked in. So it's not a perfect endpoint, but it didn't show any benefit at all, which, you you look at that and you think,

Anish Koka: was related to that. It came on the heels of that, I know is because you know, some whatever. But anyway, there was a bunch of hubbub because of that. seems like once you get and by the way, ⁓ at ⁓ right? Duchenne's, The exon skipping therapy that doesn't work, that's still on the market, right? And are still getting it. And we haven't gotten to the other

Adu: Well, the drug doesn't work, but then they continue the trial on because after one year patients can switch from the placebo, the sham arm into getting a drug arm.

Anish Koka: some of the other unpleasant crux of this is that, well, okay, society then has to pay for these drugs that, you know, ⁓ somebody to entrust somebody with saying, hey, this is the nation's person. what, what, like, what, is the best use of those resources? And if, you know, and we, ⁓ the public, at least people like me are saying there's a bunch of red flags, then Right. So the one reason why you can convince people to do a trial where, you know, a trial like this is to say that, you will all get the drug. So ⁓ one year, everyone crosses over and gets the, ⁓ the, ⁓ you know, gets the actual active. ⁓ this was a sham control, right? sham control in this particular case, sham control was ⁓ a underneath the scalp. so, you know, not ⁓ drilling in to those red flags should carry the day. That doesn't mean that if there's something that's highly plausible that you don't do it. So accelerated approval in my mind is for things that don't have 12 red flags. And certainly the Duchenne story with Sarepta and Unicure with AMT-130, there's a bunch of red flags that suggest harm, that there's potential harm to patients and there's potentially no efficacy to patients with Unicure. you know, deliver something into the brain, have to, have to get through the skull. So that typically requires a burr hole. So you have to actually ⁓ you know, a certain size hole in the skull. ⁓ that allows you to then inject something into the deep brain. So ⁓ in particular case, the sham control that was done ⁓ in or whatnot was, ⁓ you the sham ⁓ was a burr hole. It was a nick underneath the scalp. And, ⁓ ⁓ again, ⁓ you at the curves,

Adu: What would be the harm to patients?

Anish Koka: because a 12 hour surgery where you're doing burr holes and doing a deep brain injection. ⁓ definitely harm that can happen. I mean, the same reason that these are losing their minds about, my God, you're doing a sham surgery, sham surgery, ⁓ you're a sham surgery, you're forcing us to do a sham surgery. There's a risk to doing sham surgery. ⁓ Well, there's risk to doing the sham surgery and ⁓ injecting something. that I've looked at, this is data companies released it. You look at placebo, look at low dose, high dose, you know, ⁓ not activity at ⁓ one but, but, ⁓ and then what about, what about the surrogate? So there's a couple of surrogates. ⁓ There surrogate called MHTT. That was a surrogate that was thought to be very meaningful, but ⁓ you know, tell me the, tell me about the prior trial and the other that looked at MHTT as a surrogate.

Adu: I fall on the opposite side. I think that this procedure is actually relatively safe compared to the disease itself. But this is where it's like a couple of things is that, Sarepta is probably the worst example of this. They're like, you know, number one, but there's other examples of this where it worked out okay or differently. think like going back to the Amilics example, they had a trial for ALS. They got full approval, quote unquote, promised to take the drug off the market if the other phase three failed. Yeah.

Anish Koka: that looked promising, but then when you looked at clinical endpoints to try to validate that surrogate, what happened?

Adu: Yeah, actually the original trial, the sham arm, they did have sham burr holes. It wasn't just a nick in the scalp because did want to blind it. Like if you go back ⁓ they discussed some of the trial protocols, like 2019, this is a long time ago now, but they did blind it with not full burr holes, but they did do some, like it wasn't just like a nick in a 30 minute procedure. ⁓ And then it failed and then they did. They were extremely, I would say extremely high integrity people who saw the drug fail, sold about a billion a year almost within like a few years. the sales, even though the drug technically didn't work, was pretty high. But then as soon as the trial failed, they pulled it off the market within a couple of years, which is I think what the golden example of this should have been for Sarepta. Like you shouldn't have been able to run a confirmatory trial for 10 years.

Anish Koka: This is the safety and proof of concept study of the AMT-130, clinicaltrials.gov. And this was study start date was September 2019. I guess there's another one then, because this one, this one is not, yeah, it specifically this is the 2019 sham. think this is, there's only, well, ⁓ yeah, you tell me, the but. ⁓ Mm-hmm.

Adu: There's a couple.

Anish Koka: But that's the thing, your construct is basically no regulation. Your construct is then like, what is the point of regulators? Your construct is basically saying, hope the company is a good guy and they'll voluntarily pull the stuff off because ⁓ forbid a regulator tries to get involved because the regulator tries to get involved, they're gone. ⁓ This particular one, know, it specifically says imitation SAM surgery simulated surgical procedure with skin incisions only no interest or idle injections and no burholes for the skull because the thing is you're not gonna know whether or not I mean if like if I took you and you know sedated you and You woke up with the you know with a neck. Well, sorry with the skin incision and your scalp You would have no idea whether or not you had burholes or not

Adu: So do you believe that the slowing out to three years has any signal at all?

Anish Koka: I have no idea. have no idea. somebody, look again, as somebody who, at, look, I mean, ⁓ concerns I have just looking at the data is that, ⁓ mean, okay, the historical control, how good is it? It's patients and you have the first year looks nothing like the ⁓ actual So, and then have...

Adu: I was reading through their transcripts actually like 2022 they said They're anesthesia and they're getting a little burr hole in the skull But they're not actually going into the brain perhaps the skin incision includes ⁓ both skin incision and also like so I think it's like a millimeter burr hole Just to simulate like your head hurts ⁓ Like would you would know whether or not your head hurts ⁓ If you got drilled into the skull, so I'm sure

Anish Koka: Huntington's being a heterogeneous disease, it going up and down, you have the suspicious looking straight line down. So every the benefit is all relative to the fact that the historical control arm looks as bad as it does. And so, yeah, I would say, you know, there's concern. Also, the other issue is like, go back to the drawing board, because we I mean, we need another surrogate. Like, is it possible to get another surrogate in this thing? so, you know, Hmm. I see, Okay. Maybe maybe it was a maybe it was like a start of something or what not. Because you're right on the this this study must be the same one. It says no burr holes through the skull. So maybe they did something at the surface of the skull or whatnot. OK. probably somewhat academic. So no. Yeah. Not a burr hole is actually going all the way through and carving out a little hole.

Adu: I'm sure that happened. Yeah. Could that circuit be NFL, the neurofilament light chain? Yeah, exactly.

Anish Koka: Well, that's the question to ask you. How has that been validated? Is that able to be validated in any way?

Adu: this point is more validated than the mutant Huntington's would have been. Even now, I would say like some of the cohort studies that they've done. And in terms of the external control, there's 12 patients on drug, but they do match to 500 plus patients in the external control arm. So it's not 12 to 12. It is, they're matching to a larger cohort, which is why it looks so uniform, obviously. If it was just 12, it'd look a lot more like this. I the fundamental issue with Unicor putting aside all this discussion is that you can't flip back and forth. Yeah.

Anish Koka: in the skull, is somewhat of a significant thing. So in this case, whether they started something or not. so this particular trial, nothing happened. Walk me back to ⁓ and the other company, Jidentek or Roche, or who did it? Yeah.

Adu: Yeah, the mutant Huntington's, yeah. Yeah, Roche, Genentech, same thing, Toma-Nursen, ⁓ previously. Yeah, it looked great because it reduced this Huntington's protein by quite a bit. Then it completely failed. And actually if you like numerically did worse than placebo in terms of the decline on functional endpoints. And that kind of shed a little bit of cold water on this mutant Huntington's thesis, which is like, is this actually a biomarker?

Anish Koka: Right. Right. Yeah. Another promising, another promising, Yeah.

Adu: Once you have agreement, like you can't say in June, we have agreement. And then in November say, actually, what you did is incorrect. You should, because then you're wasting everybody's time. Like you go to the FDA in June and you say, this is the gripe, at least of the biotech community, community, everyone will have this gripe, not just the people who think that Unicure drugs work. because you can go back and forth on whether it works. mean, whatever, but the main gripe is like, you can't ⁓ it's just like the actual like.

Anish Koka: Yeah. Right. Right. Right. So basically. Right, right.

Adu: And there's a second piece of evidence that there's a paper out there that collected all the natural histories of progression of Huntington's and NFL progression was actually a better predictor than mutant Huntington's disease or mutant Huntington protein as a marker for progression. So I don't think we really know the answer to this because like scientifically you'd think that toxic protein, a better marker, reduce that, win. But in practice that really hasn't worked like that. way you do it rather than what is done.

Anish Koka: Don't you know, I think that's the strongest, right. I think that's the strongest argument to be made in that, you know, if you have a company that's run, you know, that's invested all this money and now, you've switched midstream. ⁓ seems to be an issue now. So I think that, I think that's a, that's a relatively valid, valid thing to kind of say is that look, you know, there needs to be some type of, you can't be when you're, when you're that. So, ⁓ Right, right. So, so I say that, again, it just shows you the weakness and the difficulty, mean, in kind of difficult diseases. And Huntington's is really difficult disease. mean, know, heterogenous, different people have onset at different ages, you know, ⁓ so you may diagnosed with it because of course the diagnosis is relatively straightforward. You have a history, plus you have, you know, It's just, you know, I think one of the issues is that have is that, you know, part of the reason for all this concern is that there's a lot of folks that are, you know, invested in Unicure, right? And they've invested because the FDA said something. And so it's not so much that, my God, the company, ⁓ my God, the poor hunting disease patients. It's like, ⁓ I've just, I've just invested millions of dollars into, ⁓ this thing based on what I think, you know, Unicure, what the FDA is going to do. And now the FDA is flipping that. That seems to be more driven by, you know, gene or not, but then in terms of when you get it, you know, that becomes a little more complicated. And there's some suggestion of like how you can predict that and stuff, but still not, not, not straightforward, which is why, you know, it's, you know, their disease process and their, and their disease processes and some disease processes, as you know, not no doubt already know, are just inexorably progressive, right? You're just constantly going down. I mean, it's just like, it's, it's, it's crazy to watch. And like, if you watched and there are certain diseases that do that. financial gains in the biotech community than necessarily that ⁓ company has to raise or ⁓ rate and all that other stuff. Is the fact that there's such a large number of analysts that are covering this that also own the stock in the company, how the heck are they going to make any objective commentary on the regulator ⁓ But then there are diseases that you have this kind of, know, and multiple sclerosis is one. mean, there's different subtypes of multiple sclerosis based on what type you have, whether you have the one that's just consisting of that or not, you know, different heart failures are like that as well. Anyway, it's complicated. the, you ⁓ that So MHCT, which now, you know, based on Roche and Ententri, basically is not ⁓ that great a surrogate.

Adu: I think that objective commentary is probably hard to come by because people do seem to be a little bit divided here. I would say that even if they aren't making objective commentary, they are making, I would say, logical commentary, which is how do know where to invest if you don't know how to run the trial? And if you don't know what And maybe Unicure is in the wrong here. There is definitely some technical aspects of this, of like how they edited the stats plan, maybe the FDA didn't like, how are they going to run the confirmatory trial? How does it actually look like on the surrogate? Like there's a lot of different things you can ask. The thing is, ultimately, if you zoom out and you look at it, FDA agreed to something, FDA didn't agree to something. And now overall, people do have to put their money towards companies to invest in drugs.

Anish Koka: What does Unicure do next? Why do they not abandon AMT 130 after that sham trial looks so bad?

Adu: Essentially, if you extend the, I would say that most investors had pretty much left this for dead. I would say like as a company, if you zoomed out and looked at the stock price what the perception is, like if you use that as a proxy for the perception of the drug, it was really difficult to, people think that the drug worked, but then they come out with two year data and the two year data from the trial. ⁓ And that money that could have gone to invest in Huntington's disease, maybe they now don't want to invest as much. Like, I think that's the gripe, right? And then the other gripe is like the politics of like going on the CNBC and that I'll put that aside at the end of this, like in March.

Anish Koka: Mm-hmm. No, I think that's a- Right.

Adu: shows that even though the one year sham controlled effort didn't work compared to, you know, propensity matched natural history, we do have ⁓ lot of natural history in Huntington's. And if you match for specific characteristics, the data actually looks quite good. Like they quote 80 % slowing. I think that the percentages get a little bit difficult when you go from like minus 0.2 to minus one, just because percentage, maybe not the right proxy for absolute Delta. But there was a, I would say like

Anish Koka: ⁓ right, right, right, right. Yeah, right. So, right. But of course, what does the what is the public trading of Unicure and Unicure stock going up and down? What does that have to do with money within the company? Meaning, like, do we have what is the point ⁓ this? Yeah, the point of having a publicly? you investing or other people investing or mean, Dirk Haseker, who was incredibly upset by this, who, know, three three quarters the way down his blog, where he's going through this mechanistic plausibility of what may work. He says, you know, I should at this point disclose the position that and disclose that I have ⁓ a position in Unicure. It's like, okay. what, what, I mean, what is the, what's the value society of, of that happening?

Adu: maybe a signal there. subsequently, after that signal, the FDA agrees that maybe it is ⁓ possible them to file, if a three-year file for approval based on like an accelerated pathway, a potential accelerated pathway, if the three-year results are great. mean, and biotech companies know that they go public so they can raise money to fund trials. That's basically it, right? And a higher stock price.

Anish Koka: Yeah, you'll to tell me, do you have that picture, by the way? Do you have the slide of the historical control comparison? Do you have that by chance? Right. But you've raised your money. Right. OK. Sorry. Go ahead. So you mean your repeat your next series of funding basically. OK.

Adu: So how are going to fund the Phase 3? Like how are you going to fund the confirmatory trial? Maybe, don't know, Unicure probably has enough cash to fund whatever Phase 3 they need to. multiple, let's see, can pull it up. You you raise at $70, you need to raise 10 % dilution versus a 20, you need to raise 30 % dilution, like, whatever you obviously like the amount of shares in the company changes and depending on like how much money you need to raise to invest in said company. And then you invest in like, that's the whole point of going public is to provide a more liquid market for these like weird asynchronous events. And I mean, like we could go into like, you know, why financial markets exist, which I don't think that's the point of this podcast. You can, there's a few from back in the day. They have a bunch of different, I would say, data cuts here. think the main one is the two year one and then the three year one are probably the most important ones. They presented it in June of 2024, believe, or July of 2024.

Anish Koka: Yeah, if you have it readily available, just let me know. think I can, are you allowed to share your screen? You can share your screen too. I had it, I of course was tweeting a bunch about this a while ago and I tweeted that slide. But we about it, the Unicure, I'm historically matched perspective controls Yeah, no. mean, so that's, that's again, that's a good point. I mean, you know, so it's in terms of that next series of funding it's almost always going to be somewhat dependent on stock prices, ⁓ the stock I guess. And so, so that's what, ⁓ so it's

Adu: Yeah.

Anish Koka: matters though again you gotta understand looking from the outside from folks that aren't invested in are just looking at the data to say like hey does this thing work or not right? i mean that's i spend a lot of my day or not a lot of my week i should say you know trying to figure out whether ⁓ drug x works for a patient correct you to understand that like you know don't care what what's happening to the stock price It's challenging because there's tons of that you don't measure for. so do you really get the same, you're trying simulate a sham control. certainly the, look, you shared something. ⁓ Did you do ⁓

Adu: Yeah, I think I just, sent over the July 2024 results. I think it's slide 16, call it.

Anish Koka: of the company and that's not influencing my decision. I'm making a decision based on my evaluation of my critical appraisal of the trial. And so it is extremely weird for us in the clinical community to watch cure, you know, the, the, the biotech community that's long on whatever ⁓ has ⁓ analyses that are very deep, but are ⁓ incredibly motivated and, you know, clearly ⁓ ⁓ you know, based on the, ⁓ based on the, ⁓ ⁓ I see. Okay.

Adu: that has the external control on it.

Anish Koka: Okay, great. Let's see if I can share this screen here. Alright, so we're looking at this. Alright, so this is the... on their ⁓ interest ⁓ it. Some of it is extremely good ⁓ of course, but ⁓ gets very hard to untangle, especially when they have a on how regulation should happen, because obviously if regulation happens one way, their position is worth X, and if it happens another way, their position is worth ⁓ negative 4X something like that. So it's interesting.

Adu: This is the external control. I can pull up the...

Anish Koka: this is. Oh, OK, OK, OK. This is the external control. Yeah, there you go. Yeah, so the external control is just like this straight line down, when interestingly enough, at least for the first year when you looked at the sham, which was basically, that's your control. That's your, the external control, the troubling thing that I wanted to bring out was that the external control, certainly in the first year, looked nothing like the sham control in the first year.

Adu: placebo data as well.

Anish Koka: What are your what are your you're ⁓ your world, ⁓ in your, you opinion, the the best here would have been to look, ⁓ this done, it, move it along. Is that is that the general what was done was commit to it. And then if you're making a new decision, you know, on something new that's coming on. ⁓ So yeah, so mean, how many patients in this external control? 12 people or something like that.

Adu: Yeah, I think it was 13 at the ultimate three years. I mean, call it between 10 and 20. It's been frankly the margin of error.

Anish Koka: Right, right. mean, come on. mean, you're talking about drilling, drilling into somebody's skull, you know, for, for, you know, the shambles, but one year that was negative. They kept, kept it going. You know, they, they obviously followed these patients out and then they came up with a mass historical control ⁓ difference in said, Hey, look, compared to this mass historical control. And that's 12 patients. That's where, that's where the, no, that's how we are pretty convinced that it has efficacy and then can change course. Because the question that again folks like me have ⁓ it like we were on a path that was like, you know, if you don't do what the Peter Marks, maybe a works, let's just try it. Let's throw ⁓ spaghetti at wall and see what sticks. Anyone who tried get in the way of that was going to face serious problems. Then saying instead of saying, then the FDA, instead of saying this is all happening again in a pre, know, McCarrie pre-Prasad, you know, pre-Trump FDA, ⁓ somebody at the FDA says, gives them a green light and says, well, this is convincing or this is, you know, reasonably sounding and gives them a green, gives them the green light, gives them the go ahead. Is that right?

Adu: Yeah. I think, what was the question there?

Anish Koka: So, ⁓ were saying that it's a consistency thing, right? Where here you had a decision that was made on Unicure ⁓ the problem was changing it. But ⁓ an FDA that came and said that that was attempting to change the Peter Marks culture of, we're not gonna just approve things based on maybe could work, ⁓ especially there's a bunch of red flags. It really felt like anyone who tried to change the Peter Marks culture

Adu: There's two pieces to that. One is the original FDA alignment happened. Prasad was in 2025. Yeah, happened right before Prasad and McCary joined. So it was December 2024. And then they met with the FDA again in June of 2025, who basically, and then they basically confirmed the SAP, the Statistical Analysis Plan. And that was after McCary and Prasad was in, were in, and that was in June of 2025.

Anish Koka: Yes. going to run into a problem whether or not even they left all the Peter Marks decisions alone.

Adu: think the answer to how would you go about changing this would have been in June when they met the FDA say, actually this trial doesn't work. I think that's like part of it. Like, you know, you met in December, then you, meet in June, according to the company and say that there's also other ways you could have done this. If you want it to be more subtle about it, like ⁓ the accelerated approval and confirmatory going to look like? Like they. So the FDA had, they had met with the FDA a couple of times and in theory they had agreed both times to using these external controls as a potential registrational ⁓ effort.

Anish Koka: So, I see. Right. so here you have it here, here, just, just, just to highlight here, you know, the control arm, you know, like the actual experimental control arm, meaning, you know, what actually happened to folks is this, which looks completely different than the historical control over the first year. So, I mean, yeah, I mean, I mean, it's an, to me, it's an inexplicable decision by the FDA. I mean, again, I'm, don't follow this stuff, you know, I'm

Adu: CURE doesn't have a randomized confirmatory going on. How come? Like, what's going on there? You could have also cited some issues with the statistical analysis or ⁓ way they match the cohorts. The cohorts that they matched to didn't include brain volume. Unicure says it doesn't matter. Maybe we need to see all that data analysis and the FDA requires that instead of the requirements that Unicure showed. The statistical analysis plan can't be re-can't be changed.

Anish Koka: buried in cardiology stuff, it's hard enough to track that.

Adu: Once the trial starts, they did make a slight, just a small tweak to it. Maybe cite that instead of just like the pure efficacy. So there's probably a more subtle way to like call out the red flags or say that actually mutant Huntington's isn't good anymore. You know, back in whenever, there are other ways to do it. If you want it to go about changing that culture of the F like changing the culture of how, what, what is the bar for evidence that is required? I think ultimately somewhere in the chain of like the company decides to stop the trial because of lack of efficacy. The FDA decides to reject the trial because of lack of efficacy. The doctor decides to not give it to the patient because of lack of efficacy or the patient decides I don't want this because it doesn't work. Somewhere in that chain someone has to make a decision. And ultimately more on the side of the doctor and the insurance company making this decision than the FDA. Just because I that the FDA is great.

Anish Koka: Hmm.

Adu: I think we do a great job. We're the best drug regulator in the world. But ultimately, obviously I'm biased on this. I'm in med school. I think that clinicians are probably would in an ideal world would be the best arbitrators. Right now, you can argue that maybe they're more biased because they don't interpret the data as deeply as the FDA would. They don't see everything because they're so busy in their day to day lives. So they just see a successful drug. They assume once it's approved, it's effective and safe. And there's puts and takes to that. I think in an ideal world, it's like an insurance company and a clinician decision to make the decision on whether or not to pay for and approve the drug because those people probably have the most incentive for cost-effective, patient-effective care.

Anish Koka: So what do see the role of the FDA to be? What should the FDA's role be?

Adu: think the FDA role changes depending on the the asset and indication, but ultimately it's just to approve safe and effective drugs. And however that means, it will depend heavily on the indication. And it's made, I think that I swing more on the safe than the effective. I think that you, in the current world, you definitely need the FDA to judge efficacy as well. And obviously, FDA goes in cycles. Like, pre-1962 enthylidomide, the FDA was a lot more hands-off.

Anish Koka: Yeah, so that's the key Favre amendment that started this whole part about the FDA now labeling things as effective or not to try to protect the ⁓ American public from snake oil. So it went from primarily a safety focused arm to ⁓ looking safe and effective. And ⁓ of yeah, go ahead.

Adu: Yeah, the Kefauver-Harris Amendment, like, start this. Exactly. Exactly. And then the HIV AIDS epidemic in the ⁓ and 90s, switched to tonight you introduced accelerator approval ⁓ then oncology was too strict. And then Pasder, ⁓ had some personal things that happened to him that he became a little bit more flexible on PFS as an approvable endpoint. And then, you know, everything goes in waves. CAR T, a bunch of it ⁓ approved on PFS. And now the FDA is going to require randomized control trials for these blood cancers.

Anish Koka: Yeah.

Adu: Peter Marks applies, approves a bunch of rare disease programs and now the FDA is swinging back a little bit and pushing back on maybe all of these shouldn't be approved, right? This randomized withdrawal trial doesn't work. You need a little bit more randomized control data. it goes in circles and there's going to be difficulties when these things change. And just, yeah, I think ultimately the FDA's job should be some efficacy, though I would like in an ideal world that committee at the FDA doesn't decide the efficacy for a drug for everybody in the U.S. I would rather have a clinician who can appraise data, each clinician that can appraise data make that decision for the patient.

Anish Koka: Right. Yeah, right. I totally agree. think this whole business to me seems like there probably shouldn't be some grand person or team at the FDA. Now, at the FDA, that's making the decision for all Americans. And yeah, let the chips fall where they may. But I think, yeah, I don't know. don't know if that's a feasible goal. The other very interesting thing when it comes to Huntington's that I'd be remiss not to mention is the fact that there's of course a lot of activity in the space, right? There are other trials that are ongoing, correct? And there's some actually very promising stuff that don't require 12 hour surgeries with burr holes and brain injections. Is that correct?

Adu: Yeah, Skyhawk Therapeutics has some really promising, I'd say six month results, which actually show improvement, which is what you'd want to see, right? Like if you had a randomized controlled trial, this goes back to the original, one of the original trials for Spinraza, which is for ⁓ SMA, which is like a deadly child, deadly condition in children. They had to run an RCT. Which is crazy, right? Like people who don't want an RCT for a disease of 40 year olds, but then if you go back to history, the FDA wanted an RCT for a disease of two year olds that would die. They ran the trial after six months. I think it was six months, maybe it was a year. The drug was so good, they had to stop the trial for equipoise that you you basically had to stop the trial early and say, well, we just got to cross over everybody. If you ran a randomized control trial like Skyhawk and you saw that same improvement on drug, you would just stop the trial early. That's I think an an impressive drug. There's another one from PTC, looks a little bit less impressive, but if you use a criteria of like NFL and a slowing on this functional score, it probably has an approvable data set at this point.

Anish Koka: No, so these are all sham, mean, sorry, placebo sham, placebo controlled Yeah.

Adu: Yeah, these all include placebo or controlled. Yeah, they all include placebo or champlate, whatever you want to call controlled data sets anywhere from a year to 36 months, depending on the severity of disease.

Anish Koka: So, if you have all these other stuff in the pipeline, who cares if like, don't care. mean, it's not even this business of like, we have nothing else for them. It's like, okay, why don't you just, I mean, maybe the smart thing politically to do would have been to, you know, while you were saying this being like, hey, we have these drugs that are coming in, we're gonna get them on market in like a year and they're way better or they're much more likely to work. Would that have been a better approach?

Adu: to say, wait for these other drugs.

Anish Koka: Well, say that, look, looked at all the data and we have huge about this being likely not in terms of Unicure. But ⁓ hey, hold folks. we have got a number of other very promising things that we are going to, that we are likely to approve based on whatever.

Adu: Yeah, I mean, the FDA, if they were a political agency, they'd certainly be calling that out and they'd be rushing to talk about these other drugs that look fairly promising. Unicure has somehow turned into like this keystone of like the unmet need at the FDA and like this whole placebo controlled data, when in reality, like it didn't seem like anybody cared until this three year data came out and reversal happened. Like I think that if Unicure in June ⁓ had with the FDA and they said this data isn't good enough, like... none of this would have happened. Like none of us would have really like paid that. Like I would have been paying attention because like I said, somehow I'm addicted to biotech. But like most people in the world probably wouldn't have heard of them. I think that's not the FDA's role to call attention to promising therapies. Their role is to regulate them. ⁓ It is the biotech committee and it's our.

Anish Koka: Yeah, so that June 2025 meeting, Prasad is, well, I don't know if Prasad's in the meeting, but clearly it's Prasad's FDA. So the question is, yeah, it'd be interesting to know what exactly happened in that June 2025 meeting. Is it that Unicure is pulling a fast one here or is that actually what happened?

Adu: Yeah, that's certainly a, uh, it's we'll see. think that that's, I'm pretty sure at one point somebody is going to file a lawsuit and then it'll be fun to put that into a GPT project and see what actually happened at the end of this. Um, yeah.

Anish Koka: You So you actually, I from an standpoint, it makes a lot of sense to know a lot about your regulator, right? So you spend a lot of time going through Prasad's.

Adu: spent probably a little too much time given that he's now out at the FDA going through his entire podcast history and pulling out takeaways from it, like he likes to regulate things and whatnot.

Anish Koka: Yeah. Well, what do you think about it in general?

Adu: Um, I think he's Somewhat. I don't think, I think he thinks he's right in every situation. And even when he may be wrong, I think the Moderna situation is probably the best example of this, which is like, if he had just rejected it, it would have been a lot less than if he had a refusal to file. Like it makes no sense for that case to be a refusal to file, but there are some like, maybe we don't give full approval to this and like considerations that you can talk about with the drug itself. It's more like, think that he wants to be judged, he wanted to be judged during an execution at the FDA and that just didn't fly. I mean, he's an opinionated person and opinionated people, there are other people who take the other sides ⁓ those arguments in every single case. So anybody who's opinionated and disagreeable will obviously get into arguments. So I'm not surprised by he was a controversial figure. I do, like I said, I think that

Anish Koka: Yeah. Yeah, yeah, yeah.

Adu: He can sometimes miss the force for the trees. So way he did things ultimately wasn't the right way ⁓ how the FDA would operate.

Anish Koka: So by that you mean like Moderna in terms of the process of like, right.

Adu: Yeah. mean, the Moderna, like, refusal of file is typically like your drug, like you did something really bad. Like there's no reason, like you should be able to file for approval. And then if you don't think the merits, the drug warrants approval or full approval, you know, go, maybe you don't approve it. Or like, maybe you give accelerated approval condition conditional on a full trial or something like that. But the refusal of file is like, why?

Anish Koka: Yeah. Yeah. Yeah. Yeah. It's. Yeah. It's so it's so interesting ⁓ that, again, just a clinician who ⁓ appraises evidence and, you know, sees a lot of patients on a daily basis. so, you know, my eyes start to glaze over when when, you you ⁓ to talk about like, you know, like, so stat news, you know, this is this is the stat news thing. Stat news is all about like, ⁓ did know, RTF and ⁓ you know, this is just not process why has it's done and blah, blah, blah. So anyway, it's super, super interesting to hear the perspective on that front. There's clearly a way to do things and, you know, we'll be the, we'll be somebody who does not kind of go about things in the way that it needs to be done. What do you think about the media coverage in general? about the FDA and then science? Do you find it useful? And then do you think about, you how they covered the Prasad situation?

Adu: I it was just fine. I don't read much media anyway, so I couldn't tell you exactly how they covered it.

Anish Koka: So most of the stuff you read is primary stuff, primary data that the company gives and then other blogs and stuff like Dirk Haussecker and those type of guys.

Adu: I mean, pretty much it's mostly like SEC filings and presentations from the company and stuff like that. As far as media goes, full I am a broke med student and I do not pay for a stat new subscription. But occasionally people send me articles. I think they do a good job. think like they do a good job. think the Wall Street Journal probably like does a good job. Like ultimately these journalists, they're not allowed to take positions in whatever stocks they write about anyway. So obviously they don't have a biased opinion. Maybe they're biased a little towards sensationalism, but aren't we all? If you're, you know, writing online, you're obviously in that sphere to some sort like biased towards getting attention. I don't find any issues with the media coverage, honestly, like,

Anish Koka: You don't think like the Wall Street Journal editorial page, you know, is like a player in terms of our stat news or endpoint news. They're actually players that are attempting to, you know, decide who gets to be a regulator or not. And then there's this cozy relationship. Why is that? Well, of course, you know, there's a lot of advertising dollars that flows to stat that there's, you know, and there's a fairly comfortable relationship with the Wall Street Journal editorial page. And certainly large corporate interests and so they want ⁓ regulator that's not super, you know, difficult for, for their clients, shall we say. So, ⁓ yeah.

Adu: Yeah. Frankly, I think that Prasad would have been the best thing for them because everybody tunes in when there's drama and Prasad created drama like we're freezing a file. Now everybody's listening to stat and it's like rejected this. Now everybody's reading stat news and Wall Street. Like they just like quick. Like if you want it from like, think objectively from an attention perspective, they might actually want more attention. But then like I'm drawing lines between like the advertising dollars and this and that. And then I have to like go through this like Arnold Venture stuff or maybe like they don't want rare disease approves. And it's like, what are we doing here?

Anish Koka: Right. Right, right. Yeah. Yeah. Yeah, sure. That's fine, yeah.

Adu: Like I look at the data, I think the drug has a signal. I don't think the FDA should go back and forth. Some people think it should be approved. Some people don't. I think it should have been approved with a really fast accelerated approval trial. Just confirm whatever efficacy that it has. They don't have that. So like they don't have that trial. Therefore, I don't think it should have been approved. If they did have that trial, I think it should have been approved. And that's kind of like ultimately where I land, which is like, when you look at like the incentive structures on this,

Anish Koka: Yeah.

Adu: it gets a little too complicated for me. A little bit too, I'd say, conspiratorial for me.

Anish Koka: Mm, yeah, yeah. Yeah, I see. see. All right, sir. Well, thank you so much. you know, thought your perspective would be super important because you know, kind of represent you're probably the only ⁓ biotech person who can who's going to ⁓ come with me and not curse at me the whole time. So ⁓

Adu: Yeah, no, I think it's a reasonable conversation to have because if you think that this drug has zero signal, I think you're wrong. And if you think this drug is God gifts to Earth, then I think you're wrong. And I don't think either of those is disputable because the signal with the external controls is quite... There is a signal there. 13 patients is not nothing. And this is a progressive disease. And progressive controls has something to be said to it. I don't think it's completely inert.

Anish Koka: Yeah. Mm-hmm. Yeah. Yeah, but it's like saying, you I mean, you're, you're, you're going to see this when you're, if you become a clinician, ⁓ is that, know, like at some point you've got to make a decision or not this path or for the patient. certainly ⁓ good clinicians just try and every single thing ⁓ that may have potential of working. Right. So you have to at some point make a decision ⁓ and, and so ⁓ So if you it. ⁓

Adu: plan.

Anish Koka: it's likely that is very likely you can predict that the Exxon 51 Exxon thing did not work. All right. So, ⁓ so person who says that ⁓ is things right. And ⁓ it's like a two sides thing. Well, it's like, well, it's ⁓ a it could work. It's like, ⁓ look, a very high probability this was complete BS based on everything that we've seen. ⁓ And think, and I think there are folks that ⁓ again, going to ⁓ not be falsely humble here because, but you know, hunting disease sounds incredibly complicated and I don't want to, I don't want to base everything on like one little slide of, ⁓ you know, the sham control thing, doing what it, what it does. But as you yourself have pointed out in your sub stack, there are some pretty big red flags. So you can see why folks would come down on that and say, not ⁓ Do the trial first. That's again, just from a pure critical appraisal science perspective. You're highlighting, I think, a reasonable point of saying, hey, look, in June, said this company is doing this. The investor community thinks this. You can't just suddenly switch at that point. ⁓

Adu: Yeah, I think so some of the neurologists from CURE and even like general do think that these results are meaningful. ⁓ beyond just the investor community, the scientists do think like, hey, this is something.

Anish Koka: Yeah. Right. Right. Yeah, right, right, right. Yeah, I...

Adu: Right? So it's not just the investor community that thinks that this has a signal. The funny thing is actually, this is the part of the reason I like investing is that it's, you know, when you're right and you know when you're wrong, there's a scoreboard every day. So I think like when you, when the, when the trial works, the market will tell you because the drug is either worth more or less versus like being the thing about being a clinician is that, sometimes the drug works. Sometimes it doesn't because it worked for this one. It doesn't work for this patient.

Anish Koka: Correct, yeah. Yeah. Yeah, yeah, yeah, yeah, right, right, right. Yeah, I really think a lot of the gap here that you're seeing between even just ⁓ and ⁓ is Prasad's basically much more like me than he is like you your community, you and your community. And that isn't necessarily we're not not not saying anything bad one way or the other, per se, but it's just incredibly different perspectives because ⁓ we're in a space where, ⁓ you some of my heart failure colleagues anything that comes out that is like sounds great, has some interesting MOA, they're like, absolutely, you know, this must work. And I can tell you that, you know, recent history and ⁓ even longer than recent history is absolutely littered with things that, that, you know, ⁓ it's why, why, why did you think that could possibly work? There's no way that could work anyway. ⁓ So, know, this, certainly the community, the KOLs and stuff, ⁓ you know, always, but yeah, but it highlights the fact that, know, if obviously, you ⁓ if some folks that are experts, this happens in medical malpractice all the time, right? You have an expert that believes that care was ⁓ below standard care, and then you have a bunch of other people that disagree. It doesn't matter. One expert is all you need for the case to kind of ⁓ proceed. So I hear you. You're making good points, ⁓ but ⁓ it certainly is. complicated. the perspective is, think, incredibly important for us as clinicians to understand because we're so not, we're so far removed from the community, rare disease community and stuff. I hope ⁓ that'll good. Yeah.

Adu: Yeah, we should chat about the FFR results recently. I've seen you write and written a little bit about that. Yeah, could go on about that as well. But ⁓ I think, yeah, that pretty much wraps up the discussion on the Unicure stuff.

Anish Koka: ⁓ my god, yes, yes, yeah. Yeah, absolutely. All right, sir. Thanks so much. All right. We'll talk soon. Take care. right. Bye.

Adu: All right, see ya.