Ash Jayagopal: Closing the Clinical Research Innovation Gap

Ram Yalamanchili (00:04.654)

Hey everyone, I'm happy to have Ash Jayagopal join us today. Ash is a good friend. We've worked closely over the years and really excited to have a conversation about what I think are really relevant topics, especially now. Just coming out of the JPM conference and saw a lot of commentary around rare disease and regulatory pathways towards rare disease being accelerated.

as well as some of the comparative nature which we have to address between the US and some of the other markets out there for drug development. So I'm pretty excited. I think we have a lot of cool things to talk about today. In terms of Ash's background, he's been a prolific researcher as well as a senior executive leader in the ophthalmology space over the many years. He spent several years at Vanderbilt doing retina related research.

moved into Roche and most recently now at Opus Genetics working on ultra rare and rare diseases in the ophthalmology space. Ash, welcome. Thanks for being here.

Ash Jayagopal (01:11.739)

Thanks, Ram Thanks for having me. And it's been a pleasure working with you and Tilda on really accelerating clinical trial development in an intelligent and efficient way. Thanks for all your help at OPUS.

Ram Yalamanchili (01:25.026)

Yeah, thank you. So I will want to kick off our today's discussion with maybe just at a broad level, you describing your role in the recent past in the rare to see space and just tell us tell us in our audience about how you've seen the space evolve in your role in it, right?

Ash Jayagopal (01:44.603)

Thanks. I have been in retina drug development for almost 15 to 16 years. I'm currently the chief scientific and development officer at Opus Genetics, a company which was founded in 2021, and I was the first employee and founding chief scientific officer there.

Our goal is to create a sustainable engine for the development of retinal gene therapies for patients suffering from blinding inherited retinal diseases. And so we are working naturally, exclusively in the rare disease space. And I get asked this a lot from if we're building a drug development program in an indication with perhaps very ultra rare 500 patients, 400 patients.

we often get asked, well, at what point is your target patient population too small for commercialization? And my answer is not really straightforward, but it's more like the numbers, the patient numbers aren't necessarily the root of the problem. It's something we both talk about a lot. It's the foundation and the infrastructure for clinical development of these therapies. So we have these 500 patients, but what if we...

bolstered them by robust patient registries globally, natural history studies, centers of excellence in clinical trial sites that we've pre-identified. have expertise in treating the patients that we're seeking to develop therapies for. Maybe that has actually a higher probability of translational success than the other program at another company with 50,000 patients, but they're scattered across the globe. The diagnostic codes are

hard to interpret and no one can really easily find those patients. So Opus Genetics, where I am, we are being, we're really a patient centric company. We're founded by a patient organization in 2021, the Foundation Fighting Blindness. We really set that commercialization strategy and began with the end in mind years before we were clinical stage so that we can be successful with the ready to clinic package. And that really has strong conviction for all stakeholders. So

Ash Jayagopal (04:00.23)

For all the things being equal with science, there's a lot of great world-class science out there. I'd say Opus really pays even more attention to beginning with the end in mind on the other aspects of clinical development. And I think this is where automation and the things we're talking about are going to play a much bigger role going ahead. So it's fun to brainstorm on that with you. Thanks for having me.

Ram Yalamanchili (04:24.558)

So interestingly, I think some of the things you're pointing out are unique to your model, right? I think Opus focuses on rare disease. It seems like it's an exclusive model towards rare disease only. In your experience, what does that mean in terms of execution or building the company itself? Do you have to make certain choices which are unique because you're focused on rare versus the other companies who are, you know, non rare disease model, right?

And where would those areas of focus be if I were to like sort of contrast it between the two types of companies?

Ash Jayagopal (05:00.635)

Sure. Working in the rare disease space, we have a certain number of shots on goal. We have to de-risk those shots with strong conviction in the science. We need a strong pre-clinical package. For many of our programs, we have proof of concept data for our AAV viral vectors in large animal models that have the same mutations of inherited retinal disease patients we see in the clinic. That's one de-risking factor. Another

aspect we think about when we select programs as a rare disease company is working with centers of excellence which have strong treatment experience, as I mentioned, with the patients we're seeking to treat. They have decades of experience treating patients with our conditions. They understand potential preclinical endpoints. They understand severity, onset of progression, and that really is informative for clinical trial design.

And of course, we need strategic partners with rare disease tailored manufacturing. We really have a scaled down problem where we're trying to make small batches of very high quality, high purity, GMP grade material to treat patients with a local retinal injection. And then we need a strong clinical partners to help interface us with our clinical sites, to identify patients and treat really the right patient at the right time.

given our limited resources in the rare to see space and the intrinsic challenges.

Ram Yalamanchili (06:32.942)

Interesting. So you essentially have this dichotomy, right? If you're running, I would say, highly prevalent disease trials, then you have to scale up. And then in your case, you're really looking for the right partners who can scale down. And perhaps neither of them are potentially the most easy things to solve for. They have unique challenges on both sides, right?

Ash Jayagopal (06:54.103)

Absolutely, and in either scenario, patient identification is equally going to be a strong challenge. In the high prevalence indications, you have competitors trying to enroll patients with the same indication for their trials at the same sites. In multifactorial diseases such as diabetic retinopathy, where we've worked in the past, you have patients on five, six, maybe seven.

diabetic, anti-diabetic medications, which may confound your assessments and of treatment outcomes in your ocular specific clinical trials, for example. In ultra rare diseases, patient identification can be a challenge where we're trying to find patients with confirmed genetic test results and genetic testing is constantly evolving. So the fidelity of a genetic test from 10 years ago may not be as strong.

as a recent genetic test with higher confidence, sensitivity, and specificity. So we're all dealing with our different challenges irrespective of the scale. They're just different kinds of challenges, really, I guess.

Ram Yalamanchili (08:03.822)

Makes sense, yeah. And bringing up this patient ID, I do have a couple of questions on that. It's one of probably like the number one reason I keep hearing for trials not going as planned. I've been to several conferences, topic is probably one of the hardest debated topic. Lots of providers who come in have different solutions for this. I'm just curious, like, you

Have you seen any meaningful improvement in this particular problem? Like have solutions come across maybe in your own relevant space, right? Where you've said, hey, this is like meaningfully changing the game here over the past few years.

Ash Jayagopal (08:45.563)

For our space, when we're looking at these large data sets, for example, claims databases or EMRs, at first, at the outset when we were starting the company, we were seeing that relying on the diagnosis in these records was making it difficult to truly find the patients we're seeking to treat, especially for inherited retinal diseases. The patient chart may not indicate

the exact molecular genetic diagnosis underlying the inherited retinal disease. It may just say retinal degeneration. So the lack of specific coding can make it difficult. And that in itself was due to the fact that treatments for these inherited retinal diseases are lacking. And so I understand the physician's challenge in really having to explain to the patient that they have a specific genetic diagnosis, but there's no treatment.

for it. And so I think that could be the reason why in many cases we're not getting a chart with the specific diagnosis in there. So I think what we're getting at is these large databases are only useful if the source data that's feeding them

It has high confidence and accuracy in detail. These databases are only as useful as the source data. And so we really need to QC the data coming in and pre-specify the parameters we're looking at.

Ram Yalamanchili (10:18.51)

So maybe another way to put it is aren't registries essentially solving this problem in a very orthogonal way? Isn't the idea of a registry that, hey, there's only 500 patients who have this particular disorder, so why don't you come in and register directly into this centralized database? And we're going to promote that through any means possible so that you know about the registry being existent.

And I'm just curious, like it feels like an interesting and valid model because, there's such a small amount of people you're trying to recruit and educate them on that particular disease. And I'm just curious to hear like, you know, on one hand, I see the needle in the haystack problem where you have hundreds of millions of people records. You're trying to find 500 of them, you know, in that hundreds of millions of patient records.

Or you can say, like I'm going to mass market in some fashion with a foundation approach and then we're going to have them registered in a registry. Do you think either of these are like, or another way, do you think the registry model is the right approach? And is that the best approach which you've seen for this particular situation?

Ash Jayagopal (11:31.964)

Sure, we've benefited extensively from the availability of patient registries tailored to our inherited retinal diseases of interest. Databases such as the MyRetinaTracker registry have allowed us to identify where the patients are with a given genetic mutation with a confirmed diagnosis such that we can inform clinical trial design and plan where we should conduct our trial, for example.

We still want to improve on the quality of information coming in and there's a timeliness factor. Registries require constant maintenance, QC and updating and to really follow these patients where it's not just one interaction of a patient visit with the database but a constant update and I think we're getting there as a field.

The onset of genetic testing with the availability of the first FDA approved ocular gene therapy, Luxterna, of retaginoparvavac by SPARC has really helped us out a lot because it's really encouraged uptake by physicians of ordering genetic tests in many cases, which can be ordered at no cost to the patient when there's suspicion of a genetic cause of the retinal visual

abnormality. And so with the onset of genetic testing, we can feed these registries with very certain information, which we need to confirm the diagnosis. And so the quality of the information going into these registries has improved from our perspective because of molecular genetic testing. That's been a really important tool for us. And so that's helped us use patient registries and use it as a lever quite often.

Ram Yalamanchili (13:23.47)

So are you also saying that most of the registries which you work with have molecular diagnostics baked into them? Like that data is available as far as how the requirement registrar?

Ash Jayagopal (13:35.834)

Not all, but it's certainly improving globally. It's quite geography specific, but the molecular testing from CLIA certified labs has been taken up by all the centers of excellence where they have treatment experiencing these patients. It's almost a certainty that at the 40 global centers that see inherited retinal disease patients where they are, that a genetic test is usually ordered.

And so there's the registry and there's also the clinical site. And one of the levers we pull is to keep close interactions with these clinical sites, keep the patient coordinators up to date, really the foundation of these clinical sites that are the patient coordinators who do so much of the work, and let them know that we're looking for a specific set of mutations or a specific patient group affected with an inherited retinal disease.

And it's not uncommon in a day or two to just get a printout of here's how many patients we have with the given indication, please contact us further. We'd like to participate in your trial. And so it's been really hard to replace the direct to clinic approach. And that's an area where we could really use automation. The sites would like it. We certainly would like it. And to really develop a seamless way to reach these patients faster through

these defined centers of excellence is an opportunity our rare disease companies have because these patients aren't being seen every.

Ram Yalamanchili (15:14.062)

makes sense, yeah. And it's actually an interesting point you bring up because even though the number of patients may be only in the hundreds in the registries, it's still a lot of work to kind of maintain longitudinal, I guess, record keeping across the board, right? Yeah. And that's certainly an area where I can see bringing automation or some form of AI to kind of help you upkeep these records without too much of a burden on either the site or the patient could potentially benefit everybody in this case, right?

Is that something you're seeing being done? there technologies who you've used or who you've seen kind of made a dent in that space?

Ash Jayagopal (15:53.5)

haven't to date, I'd certainly like to see more efforts in that area. With the increasing amount of therapies available to these patients year after year, it would be great to identify patients who are first willing to be in a clinical trial or are still, you know, in many cases the registries may report a deceased subject after, you know, tracking down the identification, the identifier.

I'd like to know if the patient opted for another therapy, a retinal prosthesis versus a gene therapy, which may fit your exclusion criteria. So getting to your point, it'd be great to overlay our inclusion exclusion criteria over a registry query and come out with a high confidence printout of who we ought to consult to assess their interest in the trial.

And I think we need more effort there, but getting the QC of the registry data inbound is going to be even more important than we can automate from there.

Ram Yalamanchili (17:02.574)

Gotcha. I'm sure you've seen claims that we are using an automation tool or some kind of AI tool for recruitment or patient identification that there have been, I at least I see quite a few of the marketing when we go to conference or something like that, right? I'm just curious, like, if you were to see it, given that you're telling me that you haven't really seen a solution per se, like this actually being solved, what do you think?

they're referring to or what is it that they're actually doing, think, when you hear claims like that.

Ash Jayagopal (17:40.334)

I can't speak for the others, but I'm impressed with what I hear about the capabilities out there, especially for high prevalence chronic diseases. There are a number of tools large indication companies may have that we don't have access to. But it's all going to come down to the quality of that source data.

and I think the same issues apply, but I'm certainly open and Opus were open to leveraging these tools as they become available for all patient populations and I'd love to learn more for sure.

Ram Yalamanchili (18:17.902)

Yeah, it's actually interesting. The model of essentially contributing data directly by the patient into a registry. I feel like there's like a lot of new and upcoming tools which are showing up. For example, Chad GPT Health just came out, I think a couple of weeks ago. It's essentially a volunteer model where you can hook your Chad GPT instance into your EMR, your physician's systems and things like that, the lab record systems.

I'm kind of curious, maybe at some point we'll end up at a place where the registries as of today evolved from just saying, you know, I'm going to fill out a form and put my name down to perhaps a lot more than that, right? I mean, do you see that coming or do you think that's like a pathway which we should be promoting as an industry?

Ash Jayagopal (19:04.007)

I'm glad you mentioned that, Rahm. I mean, it hits on a point I'm really excited about is empowering the patient community. These are highly motivated patients that we're seeking to treat. They're losing their vision day by day. Many are already legally blind at the time of diagnosis or shortly after.

And these patients and their families are highly motivated to seek out any tool to inform the patient community.

the drug developers about their disease diagnosis and their availability for a clinical trial. I think if we empower the patient and make these tools available to them, I think we'll see rapid adoption. And then on the back end, we can perform quality control and clean the data to develop really a high confidence bottom up registry, which I think is a very interesting model. I'm excited to see how it evolves.

and maybe it becomes a really important tool in our arsenal sooner rather than later.

Ram Yalamanchili (20:12.494)

That makes sense. Yeah, I think there's something there. It feels like there's something interesting and high potential coming in with those type of tools and new avenues. Another related question for me is, generally when I hear not just rare disease, but any trial planning, there's a prevalence factor you look at from literature or national history studies, things like that. And then you have your planning to say,

per site per month, what is my recruitment rate? I'm sure you've done a whole bunch of that sort of planning and I can't speak for you, but many have told me those numbers generally just don't get them to the, even close to where they need to be in terms of recruitment goals, actual versus estimate. I'm curious, what do you think is the reason? know, prevalence, I would assume, is well studied, especially if the diseases are sort of well understood.

Is there an underlying scenario here which is leading to this kind of like a misplaced trust on a number during trial planning?

Ash Jayagopal (21:23.589)

Sure, that's the grand challenge is for rare disease companies like ours to master their patient prevalence as early as possible in the development process. And it's a critical number.

that really should involve multiple types of investigations. So we see the number in the literature, which could be a number of reports from a representative clinical site. It could be allele frequencies calculated from the genetics. It can be using any of seven to 10 global databases. We collaborate with a wonderful company called Genescape.

to perform population genetics modeling as well as meta-analysis, leveraging these databases. But even when you have all those tools, you still have to account for geographies with genetic mutations, founder effects, the severity of a given mutation. With some mutations, even siblings with the same affected mutation may progress at markedly different rates.

And so the prevalence doesn't always tell us the addressable prevalence, and it doesn't always tell us the addressable prevalence in the target markets or geographies that you're aiming to run your trial in, which will then feed into your recruitment rate. And so it's a complex answer, but I think any company relying on a published number for their prevalence, especially in a rare disease,

likely has more work to do to be truly successful in the clinic and getting to approval for sure.

Ram Yalamanchili (23:09.986)

I see. So that's interesting. So you're saying the actual prevalence number published may not hold true in different geographies or even in the same geography, even between siblings. So there's like, I guess, a second and third order information which is missing just by going with that one particular statistic, right?

Ash Jayagopal (23:30.779)

Correct, and only until recently, as I mentioned, did we even have widespread genetic testing to confirm the diagnosis. So many patients can get the wrong diagnosis or are misdiagnosed. And so the criteria for the diagnosis is even equally important when calculating the prevalence numbers. So that goes back to all sorts of source data going into claims databases and EMRs.

Ram Yalamanchili (23:39.427)

Mm-hmm.

Ram Yalamanchili (24:00.974)

I see. So apart from finding patients to be able to run trials and prove their efficacy, recently I've been reading and following the current FDA leadership. I think Mercari just had an interview come out of JPM this year. He's talking about new type of trial designs being allowed in specifically rare diseases. think cell and gene therapy was another particularly interesting topic they brought up.

I think Bayesian trial design is another one that seems to be a lot of hits right now on the news. So I'm just curious, what's your thoughts on some of the challenges which existed and how do you see some of the new policies or least a new focus affecting where things go?

Ash Jayagopal (24:52.473)

We are regularly monitoring guidance from the agency, the FDA, as well as other global regulatory agencies to understand the extent to which they're flexible on rare disease drug development, especially in the cell and gene therapy space with respect to manufacturing, as well as clinical trial design. We talked about Bayesian adaptive.

as well as the non-clinical development pathway, the requirements for safety pharmacology to allow IND clearance. We closely monitor these. We're still waiting to, you know, see the industry come to a consensus. But OPUS, as we're, for example, FDA-funded for our Phase 1-2 clinical trial, enjoy close collaborations with multiple agencies who are really, and we're really in it together.

to accelerate drug development. I'm excited about the openness to any mechanism to accelerate safe gene therapies for patients with blinding eye diseases. And I just look forward to that collaboration. But I'm sure there's a lot more to come in that space and we'll be monitoring it closely.

Ram Yalamanchili (26:11.726)

Actually, you brought up something which I think is another topic I keep hearing, which is the manufacturing side of things, right? So can you tell us a little bit more about the type of challenges or areas where you think there's, I guess, like more work to be done to make trials more efficient and just drug development in general better, faster?

Ash Jayagopal (26:36.645)

Sure, with respect to manufacturing, I've been really pleased by the recent guidance with respect to gene therapy agents, which allows more rare disease tailored manufacturing processes to take place throughout the life cycle of drug development. For example,

I think the agency realizes that the quantity of gene therapy agent that we're making in one or two batches could very well be enough to treat the entire population living with that disease, at which point after several years, we might just revert to treating live births and incidents.

So therefore, we're really pleased with the flexibility with respect to manufacturing requirements. Just make the batches you need at high quality with no shortcuts, but still high quality in order to generate the commercial batch needed to ultimately market to patients. I'm really pleased with this.

Ram Yalamanchili (27:38.348)

What was the contrast trash? If you may, what was it like before and what is the new paradigm which we're going towards?

Ash Jayagopal (27:47.524)

I would say before there seemed to be higher stringency with more what we call PPQ commercial grade batches needed to demonstrate safety, identity, purity, and potency needed to demonstrate a process that can be reproduced over multiple batches to ensure high quality and safety for the patient.

I think what we've learned is over the last two decades, manufacturing of AAV gene therapy has really evolved such that we're achieving routinely high purity batches of contaminant-free AAVs. And as the quality of the process and the stringency of our analytical methods has improved, I think the agency has also realized that the process has evolved to the point where

a higher standard now exists for even the first batch of AAV a sponsor may make. And so they ultimately are requiring fewer total batches needed to treat the patient. And we're excited about that possibility. Although every sponsor's drug and complexity may vary, we're very pleased with the regulatory flexibility since our AAVs, using our drug products,

all pretty much have clinical precedent and they've been manufactured before and used in clinical trials before. And given our tried and true AAV approach, we've largely de-risked the manufacturing quality aspect of our approach. And so hopefully we can leverage a lot of pathways that the agency is offering to sponsors like us.

Ram Yalamanchili (29:35.022)

I see. So that's the way I think about it and coming from a non-expert, non-manufacturing background is circumstances have changed, I suppose, right? Right now we have better technology, better types of manufacturing techniques perhaps. And the agency is really just recognizing the fact that we do have better technology right now. And so the standards have to move based on what the current standard of advancement it is.

Is that kind of a good way to sum up how they're thinking about it what you're excited about?

Ash Jayagopal (30:08.439)

Absolutely, I'm really happy that the regulatory bodies have always sought to consistently adapt to technologies.

to ensure an accelerated pathway to patients without compromising quality or safety. And we're really well poised to grow with them. As I mentioned earlier, we are funded by the Office of Orphan Drug Products by the FDA, and we're very grateful for that funding as it allows close collaboration to really brainstorm on better ways to deliver these ultra rare therapies to patients faster. And manufacturing is just one

lever where we can do that, then you apply the non-clinical development and clinical development pathways and you have a package that really hopefully will enable the approval of more rare disease gene therapies for our patients.

Ram Yalamanchili (31:02.926)

I think the last part I would like to bring up is, I do hear quite a bit about how much better it is or faster it is to run trials or do drug development in other countries outside of US, So whether it be China or Australia, what's your view on this? I know you're based in the US, as far as I can tell, as you said, funded by the FDA.

have a very interesting vantage point on this. But I am curious in terms of like, know, how do you see this, why are we in this situation right now? And I guess the follow-up would be, how can you see the future, right? With more aligned agency at this point.

Ash Jayagopal (31:53.116)

Sure. So talking about global collaborations, we need to go to where the patients are. In almost all of our indications, there exists a geography in which our rare disease may not be so rare. And we want to be able to get our therapies to patients in need where they are. And so that requires a global collaboration with Asia, with the Middle East, with Europe.

And we are very open-minded and we always think globally into ensure access to our therapies. And so while we're US headquarters, we're routinely working with other regulatory bodies in parallel with the US bodies to understand how we can develop a drug that's suitable and available for all populations to the extent possible. And so.

We like to think of ourselves as a global company in order to treat these inherited retinal diseases. You have to think that way for sure.

Ram Yalamanchili (32:56.502)

And Ash, from a practical perspective, how are trials so much faster in some of these other geographies? Is there like a fundamental difference in the way they're approaching their trials or something else? I'm just curious, like, you know, at least in your space, what are you seeing, which would be, you know, helping them accelerate the way they are.

Ash Jayagopal (33:19.451)

Sure, we're seeing a lot of innovative mechanisms from all the regulatory bodies, especially in the area where we work in gene therapies to accelerate clinical proof of concept to then enable larger registrational studies. For example, in China, we see the investigator-initiated trial mechanism to rapidly make, to really trim timelines to that phase one proof of concept.

in patients. And then in the UK, we've seen the specials pathway used to make investigational therapies available to rare disease patients. Those are just two of the examples where when there's a serious unmet need, the regulatory body is willing to define special pathways for sponsors.

to get to that clinical stage a little bit faster, demonstrate proof of concept in a limited number of patients, but get to some clinical trial data, which we all understand would help the company be in a better position, be better poised to raise more funds to run registrational studies and IND-compliant studies. And so I'm really excited to see how that pathway evolves. It's another example of the increased flexibility

we're seeing from agencies for rare disease sponsors.

Ram Yalamanchili (34:46.734)

I see. So in some ways, the pathways and the flexibility on the regulatory agencies helps you sort of get to a point where you can start the studies faster. But it doesn't address the second part, I guess the first part we were talking about, which is the patients, finding the patients and then the prevalence and so on and so forth, right? Do you think some of the other jurisdictions, like does China have a better prevalence model? Have they studied it better in their country versus like say here? Do you have a view on like, you know,

Is there any edge which the biotechs out there have versus us or what's your thought on that?

Ash Jayagopal (35:23.213)

Like many sponsors, there's no right answer, but we tend to look at the site level, at the investigator level. It all comes down still to the fundamentals. Where are the physicians that have strong treatment experience with these patients?

where are the clinical scientists and thought leaders in the field? And that usually leads to us finding the patients that we need to be treating. And we're lucky to have a strong network of KOLs globally in all the major geographies by population so that we're more likely to find these patients. And so the regulatory flexibility is excellent, but it doesn't change the duty of the sponsor.

to really do their part to find these patients and that's through Centers of Excellence.

Ram Yalamanchili (36:16.566)

Makes sense. Well, that's really interesting. feel like I've at least enough to be able to have a conversation, right? So this is really helpful. Thanks for sharing, Ash. yeah, I would say good luck with what you're doing. We're a huge supporter of what your work entails and I really wish there's more companies like Opus solving a very

very important and interesting problem. So thank you for all the work and good to see you.

Ash Jayagopal (36:51.589)

Thanks, Ram, I've always enjoyed our collaboration. Thanks for having me on.

Ram Yalamanchili (36:55.468)

Yeah, thank you.