Seeking a Sustainable Business Developing on N-of-1 Therapies

The ability to pinpoint the underlying genetic causes of diseases and rapidly generate genetic medicines to address them has created the potential for the development of individualized therapies to treat patients with ultra-rare conditions. EveryONE Medicines is seeking to industrialize this process and scale the development of N-of-1 therapies. We spoke to Irina Antonijevic, chief medical officer of Everyone Medicines, about the company’s business model for pursuing N-of-1 therapies for people with rare diseases, how it works, and whether it can be sustainable absent a mechanism for reimbursement.

Daniel Levine: Irena, thanks for joining us.

Irina Antonijevic: Thank you very much for this opportunity to speak with you.

Daniel Levine: We’re going to talk about EveryONE Medicines, the development of n-of-1 therapies and how the company is seeking to help create a sustainable model for individualized precision therapies. Julia Vitarello is a co-founder of EveryONE Medicines, and I think the story about her efforts to develop an n-of-1 therapy for her daughter Mila, who had a form of the rare neurodegenerative condition Batten disease, will be familiar to many of our listeners. But can you remind people about her story and explain how that led to the creation of EveryONE medicines?

Irina Antonijevic: Yes, absolutely, and Julia continues to be a very adamant advocate and an incredible supporter of EveryONE Medicine’s goals and objectives and aspirations. So, I’m really, really grateful to be able to work with Julia, and it was Julia who was first introduced to us to explore whether we wanted to join her on this path to adventure and helping develop these therapies. And so, Julia had a little girl, Mila, and Mila was doing well, actually, exceptionally well. She was a very bright and curious child, but then she started to just become more clumsy as a really young little girl. And it just deteriorated from there. Then she started to lose her vision, and Julia was, of course, desperate to, first of all, try to understand what was going on with her little girl. And she went through many different hands of physicians and hospitals, and nobody could really tell her what it was. At the beginning I think she wasn’t being taken seriously. So, ultimately what happened, and I believe it was really a coincidence, there was an article about her and Mila in the New York Times, and Timothy Yu, who is a geneticist and pediatrician at Boston Children’s Hospital, his wife saw this article and said, Tim, can you do something about that? So, this is how they connected, Timothy Yu and Julia. Tim looked at Mila’s genome, and he identified the mutation that was a really bespoke mutation in one of the genes that ultimately led to this rare form of Batten disease. So, there are two mutations that are needed, but one of the mutations was a very unique mutation. And so, he set out to develop an antisense oligonucleotide, that’s a new kind of molecule that can very precisely target certain genetic mutations. He developed this for Mila, and he called it milasen. So, antisense oligonucleotides or ASOs often end with sen, so it was Mila and then sen,  milasen. And so, he had discussions with the FDA because there wasn’t a process in place to develop a therapy for one individual, so there were discussions with the FDA, and given the advanced stage of Mila—she wasn’t doing so well—so there was obviously a medical urgency. And so, they ultimately managed to develop this ASO and give it to start to treat Mila with that. These ASOs, particularly for disease that affects the brain, they have to be given through a lumbar puncture. So, there is a needle and the liquid is injected into the spinal fluid and can then also reach the brain. Mila was treated for, I believe, almost three years, or around three years. And in one individual, it’s hard to say how good the therapy was, but it seemed to at least slow down the progression. But ultimately, it could not stop the progression, and ultimately Mila died. I think this sort of put Julia and Tim, I want to say, on this mission to really make this something that a can start much earlier upon the diagnosis and that we can create a community of people developing such kinds of bespoke or individualized precision therapeutics. And Tim has since then continued to identify some of these patients with these unique mutations and has treated just a handful of patients, and some other investigators are pursuing a similar objective.

Daniel Levine: The notion of rare disease is really a policy construct. As a result, rare diseases often get spoken about as if they’re monolithic. In reality, though, there’s a bifurcation between rare diseases that provide significant economic opportunities to drug developers, conditions like cystic fibrosis, sickle cell disease, and hemophilia and ultra-rare diseases, conditions where there may be fewer than 100 known cases. As you think about the rare disease landscape, what is the need that EveryONE Medicines is trying to address?

Irina Antonijevic: EveryONE Medicines is trying to address or is seeking to address exactly these extremely rare diseases we speak about, n-of1 plus patients. So exactly to your point, I used to work at Genzyme. So obviously there are some companies—pharma and biotech companies that have created a very viable business by developing drugs for rare diseases. And the definition of rare diseases is a bit different in different countries, but in the U.S. it’s typically less than 200,000. So, it’s a number. In Europe it’s more like percentage, but nevertheless, these are numbers of patients that allow you to actually conduct maybe even two clinical trials that are controlled and randomized—controlled where you can actually do some statistics to see whether the drug works. If you go down to very small numbers, this just does not become feasible anymore. I mean, in the extreme, if you have one individual, obviously you cannot do a clinical trial. And so, EveryONE Medicines sees itself as developing medications, drugs for those populations where right now we don’t see any commercial organization that has set out to actually develop these n-of-1 plus therapeutics. And we think that the time is ripe for that because we have already seen in, in several efforts that used whole genome sequencing, that the number of unique mutations is actually larger than they might appear beforehand because, of course, we just know of one here, one there. But if we actually more systematically screen through the genome, particularly of individuals who show some symptoms, we see that many have these unique mutations or constellation of unique mutations, and not all of them can be treated with an ASO, but those that can, we think, that this is where we want to be. We want to develop what we call a process that allows us to identify the mutations, decide if they are really the cause of the disease, obviously this is important, then decide if we can correct the mutation sufficiently to expect a benefit with an antisense oligonucleotide, and then develop it and develop it quickly through a process that we are literally launching as we speak. And we think that it is the speed that will be important so that upon a diagnosis we can quickly deliver those therapies and multiple pieces have to fall in place. I don’t want to say that it is without challenges, but we think that the time is really right for that. Because once we have a diagnosis, a genetic diagnosis, and we have the technologies to develop a treatment, I feel compelled to say we can do it. So we have to do it, and we have to find a way for this to be an approvable therapy, even if it’s only for one individual and it has to be reimbursed. We have to reinstate health equity for those families that have an enormous burden, of course, emotionally, but also financially.

Daniel Levine: Julia talks about going from Mila to millions, scaling the process of what researchers were able to do and industrializing the development of n-of-1 therapies. What’s the potential for doing that?

Irina Antonijevic: Yes, so we think it is all about making quick, but also good decisions on what is the right drug candidate. And then over time, we have a certain set of hypotheses. I think they’re pretty solid hypotheses because the very small team at EveryONE Medicines is very experienced with antisense oligonucleotides and more broadly with nucleic acid therapeutics. So, we have created—we call it a funnel—how we can rapidly select those ASOs that we think are suitable to incur benefit in the individual, but also be safe enough. And since you know Julia, well, we are talking about, at least initially, we are talking about really severe diseases. So, we are talking about children who have many seizures a day or have a rapid neurodegenerative process that they lose the ability to walk, to move their hands, to speak, potentially to breathe. So, these are severe disorders. Of course, we still want this to be safe, but we don’t have the time to spend, like three or four years, optimizing it. So, we have optimized the screening process to get to the decision of which one is the good therapy. And with this, we believe that we can scale this up because the the testing is very well informed and we’ll become ever more informed as we continue to do this over time, because what we have as a really unique situation in this case of n-of-1 plus therapeutics is that the time from designing the ASO, do some tests, and then taking it into the clinic is very short. Today it’s about 12 months, maybe a little bit more, maybe a little bit less, around 12 months. And we think that we can really shorten this in dialogue with regulators. That’s an important statement I have to make. We are in communication with regulators, but obviously they have to agree to that. But then it means that we can very rapidly feed back the learnings from each ASO that we have developed and taken into an individual or just a handful of individuals where our predictions are correct, if not, we can revise what it’ll learn. So, it’s a very quick learning process. This is what we think is scalable. And really we plan to be able to address many, many patients, I mean, as the future will sort of deliver these diagnoses. That is the goal of EveryONE Medicines.

Daniel Levine: How does EveryONE Medicines work? What’s the business model? Who are EveryONE Medicines’ customers and how does it get paid?

Irina Antonijevic: At the moment, EveryONE Medicines is a biotechnology startup. We are located in Boston and we have what we call seed funding. We have very reputable and very serious investors who are really committed to make this work. They gave us seed money to sort of start building this process—identifying certain individuals, start building this process. We are a young company, so this concept of EveryONE Medicines has been around, I would say maybe about two years, but it was really very much in stealth mode. The current team, which is fully dedicated, are employed to work at EveryONE Medicines. So, we have an interim CEO who represents one of the investors as well. And then we are three full-time employees and I’m one of them. We are doing a lot of the activities virtually, but we are building also a network with investigators who see these patients. We are also working with investigators, particularly in the U.K. for your listeners, and you might be familiar with the Genome Project, the Hundred Thousand Genome Project in the U.K., which is where they have really set out to sequence whole genomes from a hundred thousand people. We have heard that they have more by now. And so, this data allows us to also identify genes that maybe are more represented in terms of their mutations in this data set than we would’ve expected. So, the idea is to use such existing data. There are some of these data sets that exist, and then we have some investigators that see these very rare children that, you know, had a diagnostic odyssey, and then land in their hands. And then once they’re doing the whole genome sequencing, this is when they can find those rare mutations. So, I would say our customers are to some extent also families who can approach us, but then also physicians. And so, that’s why we are creating this network with physicians and then using existing genome data that allow us to identify likely mutations that we will encounter in the future so that we can actually start developing ASOs maybe based on only one patient, but we expect that there will be more patients, maybe five in the next whatever 10 years. And we will have the drug ready for them as soon as they’re diagnosed. So, it’s also important to say we are global. We are based in Boston, as I said, but we do really have an outreach globally and we definitely have engaged with investigators and genomic data sets both in the U.S., a little bit in Canada, and also in Europe and the U.K.

Daniel Levine: Is the expectation that you’ll be exclusively focused on antisense oligonucleotides, or will you be exploring other modalities as well?

Irina Antonijevic: For the moment, we will focus on ASOs, but we definitely want to explore other modalities as well. The reason for starting with an ASO is that many of the rare diseases affect children, affect more than 70%. And many of those rare diseases in children affect the brain or the spinal cord, so the central nervous system. And ASOs have this ability to be taken up, particularly by neurons, but also other brain cells. And so we don’t need anything other than a very precisely targeting ASO, we don’t need a delivery vehicle. So, we think that this is a very good starting point to build this process and start developing treatments for these high unmet need diseases. But ultimately, we think that the technologies are expanding all the time, literally all the time, so we think that we will just be watching and observing the progress. And we definitely think that there could be RNA gene editing technologies, gene therapies. It’s all about delivery, so getting the drug to where it needs to go, for instance in the brain. And so, this is where ASOs are just often easy way into that. But as new technologies are available that allow us to do this for other modalities, particularly within the nucleic acid therapeutics, because they’re very precise, they can be designed in a way that they very precisely target a unique mutation. So, we definitely want to broaden out, and I just want to say, because what I maybe didn’t really fully address, is how is this going to be a business, you know? So, we think that by talking to payers or insurance companies and regulators, we can create this pass that allows us, like for any other drug, to have a drug that is ultimately reimbursed because it provides a favorable risk benefit ratio: there is a greater benefit than there is a risk. Just because it’s for one, not just a handful of individuals, there shouldn’t be, in principle, ethically speaking, why would they not be reimbursed? So, we just have to find a way to make it acceptable from also a regulatory point of view, that this is actually a drug that has shown a reasonable profile because it has been developed based on a certain process. And so, this is what we are working on, we are generating data with new ASOs, and we want to treat some individuals while we use this data also to explain to both regulatory agencies, but also insurance companies, what is at stake here and how we think we can develop these therapies that should be reimbursed. This is an ongoing discussion, but we have had discussions already, and we think that there is—not everyone, and not every insurance company and different in different countries somewhat—but there is a recognition that we have to overcome the challenge we have today and create a path. And we think we’ll be able to do that.

Daniel Levine: There are many well-known challenges to rare disease drug development. These are amplified when it comes to developing a therapy for an individual patient or an ultra-rare disease population. On the other hand, you don’t have to go through full drug development and seek an approval for therapy to actually dose a patient. What do you see as a realistic timeline from the start of a project to actually getting medicine in a patient?

Irina Antonijevic: So, at the moment, we think a realistic time frame is about 12 months. This is ambitious and, you know it can be 14 months. But it might be 11 months. So, I think that’s a realistic time frame. This does assume that we go through the sort of testing scheme that we have devised and thatwe find a good ASO. Again, we are pretty confident we will, but I would say this is a reasonable time frame. And at the moment, the FDA has issued a draft guidance specifically on the development of ASOs for these n-of-1 patients. And so, this is kind of like a guidance document to tell us what the FDA expects in terms of testing before it can go into an individual. And the other agencies, for instance, in Europe have not issued such guidances or draft guidances. This is why it’s important for us to have this conversation with them to see what they would expect. So, this process, just the way it’s laid out right now, takes about about 12 months just because there are certain elements in it that include, for instance, some animal testing that just require a certain amount of time. But we think that with the sort of informed and iterative learning from the ASO designs through the sort of in vitro or just cell models, and then in vivo testing, we can actually speed up this process. So, our goal is to start with a draft guidance that exists, but have these discussions and generate data in the next two to three years to actually show that we can speed up this process so that it would be faster, and I don’t want to say a definitive number of weeks or months, but we are pretty confident that this can be a faster process to take a new drug into an individual, particularly with a CNS disease. And we think at the beginning, we call this stage one of our platform of our process where we would actually show that we can do that, we can develop some new ASOs, we can take them into these children with CNS diseases, and we look for safety and efficacy that has to be defined in each individual, because these children have often a very unique set of symptoms. And we want to make sure that we understand, and this is where we will be working, particularly with the families, to understand what is most important for the child, for the family, what would they want to see improve, what is creating this disease burden for them. And you know, it could be something like my child can grab the spoon again and feed itself, or my child can walk maybe slowly, but can walk, so they’re different aspects. My child has fewer hospitalizations. So, there are different ways to measure that. But we would want to initially show that we measure this in a rigorous way in an individual child, so that we can say, here we have shown you that it is safe and it has some benefit. And then again, over time, we think that by identifying what are the key features of such ASOs we can speed up this process so that we don’t have to actually do any clinical testing. The vision is the drug is developed through a process and can then be given as a reimbursable therapy to the individuals for whom it was made.

Daniel Levine: Time is such a critical issue for these patients. Many of them will seek to develop treatments when a child has a progressive and degenerative condition. And in many cases, by the time a child is diagnosable, the disease can be far along. What level of understanding of a disease and what testing of a therapy is needed before dosing a patient? And what’s being done to move regulators to get growing comfort with compressing the preclinical timeline.

Irina Antonijevic: So, these are all good questions, but many questions in one question. In principle, the regulators have considered that there must be a way to accelerate drug development if there is a severe disease, but I think it also does depend on the modality of the drug. So, this is why these drug guidances from the FDA are very specific to ASOs and these very severe diseases. This is where, and Julia puts it so well because the risk of treatment is almost always less than the risk of not treatment, because if you don’t treat, the child will certainly progress and die. So, this is where the benefit risk equation is very different from a disease that maybe doesn’t lead to rapid death. This is where it is important to have a dialogue with the families and with the child to identify what the individual symptoms are. And because you don’t have a control group, you have to think of clearer ways of how to demonstrate in an individual that the therapy has some benefit. And you know, there are some symptoms that you can measure them before you start the treatment and then you can measure them on treatment. We also have, of course, access to, for instance, imaging. When we talk about brain diseases, imaging can be a useful addition to also measuring if there’s a benefit. And you know, there are also some digital tools like variable sensors that can also measure in a finer way, certain aspects, for instance of movements that are altered and then they can be linked to just more day-to-day activities that are affected if the movement isn’t that coordinated. So, this is what we would then work out in a given individual, because, you mentioned this at the beginning also, even within a rare disease, there are differences between the individuals even though they’re all classified as whatever the disease might be. With an n-of-1 they’re just representing themselves so it’s important to really understand what the symptoms are and what tools do we have to measure them. And for the safety, it’s a little bit easier, I would say, because ASOs, as a drug class, we have come to understand certain features of them in terms of safety issues that we can monitor. And so, we think that there are certain design features that reduce the risk in the first place, but we will also monitor for those because even if it’s a fatal and progressive disease, we still want to make sure that we also learn and we definitely don’t want to create more harm. But we think ASOs are a good class for that because of their precise effects and reasonably well understood side effects.

Daniel Levine: There are a number of patients who have been treated with n-of-1 therapies from existing efforts to date. My sense is it takes about $4 to $5 million to get to dosing a patient, and this usually involves a bit of goodwill and a nonprofit setting. What’s your sense of what it will take financially and are there opportunities to cut the cost through scaling what you’re doing?

Irina Antonijevic: Yes, I think that’s exactly what we think. Your costs are about right and it is very expensive, unfortunately, even with an abbreviated package. But there are, I would say, at every step of the way, there are ways to reduce the cost at least in the future. And this is where we think it is very important to have a clear set of experiments that we think are the most meaningful to select the right clinical candidate drug, but also to have this iteration of rapidly learning and have the dialogue with the regulators. Because what we think is that fewer and fewer tests should be needed in the future to identify which ASO is the best ASO and can be therefore taken rapidly into the clinic. Because as I said, we can monitor in the clinic the safety as well as the pharmacological effects. And so, by simply doing fewer tests, you will save time and you will save costs. There is one aspect of this development, which is maybe the most difficult to really dramatically reduce in terms of cost. And this is actually the manufacturing of the drug. We think that we can understand the particular biology and the effects of the drug in cellular and animal models better and better and better, and speed up this process and do fewer experiments. And again, build the sort of scalable process with rapid learnings. Now, the drug quality is important to ensure, because you want to be sure that you have a high quality drug that doesn’t have any potential harmful impurities or something. So, you need to have a certain rigor in how you actually manufacture these ASOs. Of course, these are manufactured in a relatively small scale because even if you have a handful of individuals, you don’t need so much as when you have like hundreds and thousands of patients. So, it’s always a small scale, but the quality still has to be good because what we don’t want at all is to have something that is not really reflective of the drug, but just the drug process that hasn’t been properly quality controlled. So, this is something that we still think there is some efficiencies that we can devise in this manufacturing process, but this will be something that simply will always have to stick to certain quality standards. But importantly, ASOs, as an individual is treated and because they’re not one and done treatments, but you treat again and again, this offers certain advantages. First of all, it offers the advantage that you can actually adjust the dose if you need to. You know, there’s the best of knowledge and experience. Once you start to treat a human, you will be able to see whether the dose that you assumed would be the right dose is the right dose. So with an ASO, unlike with, for instance, gene therapy or gene editing, you can actually continue to make adjustments to the dose. And of course, in the worst case, you can stop the therapy or you can increase the interval. You know, we don’t know, maybe some disorders have to be treated for some years, and then maybe you can reduce the treatment, either the frequency or the dose. We have to also be mindful of do we have to treat for the entire life? Do we have to treat for a certain amount of time? So, this might also be different for each individual, but there’s some neurodevelopmental diseases that might require an adjustment of treatment over time. So, so all of this can be done with an ASO and we think that, again, the sort of reimbursement should occur for something that creates a benefit and by doing this across different patients and across different countries with an ever faster development, every data point will give us more lessons learned. And we think that we will be able to create really a process that goes very quickly and is affordable in a way that the drug is reimbursed and therefore if it’s reimbursed, by treatment, you can get the cost that you initially invested in developing them.

Daniel Levine: You talked about reimbursement earlier, it sounded somewhat optimistic. The question of cost is inseparable from the issue of reimbursement. What will it take to get payers to recognize the value of n-of-1 therapies and pay for these medicines without having them FDA approved or having a large dossier to establish their cost effectiveness?

Irina Antonijevic: Yeah, so this is where the conversations we have had so far clearly show that there are differences in different countries. But I do also want to stress that our goal is to have a parallel dialogue, both with regulators as well as with the insurance providers so that in an ideal world, there would be a path to an approval, a rapid approval of an ASO that has followed a certain agreed upon process in being selected and tested because I think it is easier to have these discussions about reimbursement when there is a path to an approval. But I do want to say that there are compassionate use, named patients, so there is a way, even today, and you mentioned that yes, some people have already been treated with n-of-1 ASOs. So, there is a way to have this reimbursed. It’s just on a one by one basis, and we would like to really create a process that also doesn’t require every single time a discussion and negotiation with a particular insurance provider, but that there is a certain path that is established and that if certain criteria are met, this drug is reimbursed. And again, I think the advantage of a continued treatment is that it’s not about the huge upfront initial cost, but you can monitor the benefit and the risks of the treatment over time and can adjust and if you decide to do so, also discontinue the treatment and then there’s obviously no reimbursement needed anymore. So, we think that this is actually an advantage as the first N onen-0f-1 plus drug development platform to have something that can be adjusted and therefore also adjust the reimbursement.

Daniel Levine: As we’ve seen the evolution n-of-1 therapies, we’ve also been reminded that these therapies are not without risk. There was a case last year of an ASO developed to treat a child with a rare epilepsy who died from a side effect from the treatment. How do you think about the risk of these therapies? And do patient families weigh them or do they believe that these risks are not significant enough to dissuade them because of the trajectory of these diseases?

Irina Antonijevic: We are quite familiar with the case you mentioned. What I’m saying now is my truly personal view, my strong view that when we struggle to make a clear statement about efficacy in n-of-1, because we have ways to think how we can measure benefit, but it is still an n-of-1 that is not controlled. It’s the same with the safety. So, I think the statement that this was clearly linked to the drug and only to the drug is, I think, it’s not quite possible. I think it’s likely that the drug contributed to it. We think that there are certain features of that particular drug that has contributed to it. And this is why I believe that we have created at EveryONE Medicines’ criteria that allow us to select better drug candidates that we believe will have fewer of these class side effects because what we have seen—it starts very often with just more liquid in the brain. I mean, there’s obviously liquid in the brain—the cerebrospinal fluid, but with ventricular enlargement so there looks to be more space where this fluid is. And we have seen this in other indications. And so, I think this looks like a class effect and we believe that there is a way to mitigate the risk for this effect in the design of the ASOs. But we will also be monitoring this, of course, this is one of these typical class effects and this is where, again, an ASO has the advantage to be able to adjust the dose or increase the frequency or decrease the frequency of dosing if you see something where you’re not sure is this a side effect or is this a disease itself.  So, there can be adjustments made. And then ultimately the decision to start treatment and to continue treatment has to be made with the patient and the families. And I believe that the families are different for some, they might be willing to accept more risk than others, , and this is where I think it’s important to have this conversation with the families and outline the potential risks, outline the ways to monitor for that. Again, I strongly believe that we have come to learn more about ASOs in the last three to four years. And I think some of these lessons learned are very important and I think we can utilize them to make drugs that simply have a lower risk for some of these class effects. But at the same time, we monitor for them so that we can make adjustments quickly.

Daniel Levine: Irina Antonijevic, chief medical officer of EveryONE Medicines. Irina, thanks so much for your time today.

Irina Antonijevic: Thank you Daniel. Great to speak with you.

This transcript has been edited for clarity and readability.