2023年天使综合征 FAST 年会总结及 姜永辉教授现场视频
演讲文稿:
>> Okay, we’re going to invite our next panelist speakers up. They will be talking about the recent exciting news that we announced about a large NIH grant to advance a novel CRISPR based gene editing delivery platform at Yale University. They’re part of the project team along with Jennifer Panagoulias. This is really a remarkable example of cross collaboration, and in my years of working in patient advocacy and NIH, I’ve never seen a model like this. We should be really proud and I’m really excited to hear more about it in this presentation. [ Applause ] >> What an incredible honor to stand on stage not just representing our community but a drug development program. This is exactly what FAST has been trying to do since its inception and this is really a milestone that’s unbelievable for our community. I think we should be very proud of the fact that patient advocacy is actually taking things from bench side to bedside and this is just a true example of that. We are a team working together, and we’re going to talk a lot yesterday and the day before we talked about CRISPR technology so I’m not here to explain that technology to you but we want to show all of the steps of drug development. We have lots of disclosures. Here they are. You can read them. This is a graphic you’ve seen throughout the entire weekend. It’s important to understand how we can take something from early discovery through foundational NIH and other nonprofit research funding efforts and even some for profit funding efforts and go from discovery to preclinical which is proof of concept which we talked about yesterday and take this idea, put it in an animal model to show this can work and are there early signs that it’s safe. The next step is to say is this possible for humans and can we develop this drug for humans. That’s where we have to get to a human candidate. Once we get to that human candidate, we need to understand the safety profile of that which I’m going to talk about. We have to develop endpoints and biomarkers and that’s through the efforts of the ABOM that we talked about yesterday. Then of course the community effort, we need all of you. We need your enrollment in the global registry to have a good sense of where these trials can run. Finally, we need to run the trial which we are actually funding the infrastructure to create places like rush who can run these trials that are very innovative and complicated and we have the confidence to do it. The team at Yale is going to talk to you about the first part of this ecosystem which is preclinical and drug discovery.
>> Thanks, Allyson, for the great introduction. I’m a professor at Yale Medical School. A great pleasure to be here. On the team my role is engineer. So my role is to deliver the payload. In this case CRISPR to the cells to the right place which is the brain. That’s what I’m doing. I’m very happy to be here to share the journey, how collaboration between engineer with biologist to lead to this translation. What we need to deliver is called CRISPR. CRISPR actually is developed by bacteria to prevent infection. We want to borrow the system from bacteria for human use, use CRISPR to reactivate the UBE3A gene protein. For this application we want to deliver a system with two components including protein shown in the cartoon gray and [ indiscernible ] shown in orange. This is a two component system. My job is to deliver to cells in the brain. Two major challenges. Unfortunately in the virus, the best approach to deliver the CRISPR system for two reasons. First CRISPR isn’t very large, proteins. You can potentially use two virus but the more virus you use, the more potential side effects you will have. Another potential limitation, in this case we deliver bacteria as I mentioned a moment ago, so viral delivery has an advantage. So that’s good for gene replacement therapy. But this could be a problem in this case, we want to pass bacteria for many years and this could be dangerous. As an alternate approach you can potentially use nano particles which can mimic a virus. It might be new to some of you but actually you may use particles for COVID 19 vaccine. Both Pfizer and Moderna use nano particles. Then the second challenge, deliver to the brain. As the doctor mentioned this morning, the brain possesses a unique organ called blood brain barrier. The vessels in the brain are unique biologically and physically. As a result it’s very difficult to get molecules to penetrate it and get into brain tissue. One way to bypass the blood brain barrier is to inject ASO into spinal fluid. The hope is that the molecules can diffuse into the brain which means in some cases the molecules can diffuse itself. The problem is that for nano particles and true for most virus, too large for penetrating into the space. The space is typically about 10 to 13 millimeter. Payload to be delivered called CRISPR, we can do chemical modification. That means we can [ indiscernible ]. The beauty of this approach is that after engineering the result in [ indiscernible ] is very small. In terms of size, perfect for diffusion in the brain making this particularly suitable for brain penetration. Of course we need to test these particles in cell culture in vitro and in vivo. In the left panel we found that with single treatment you can essentially in almost all cells. We also test in mouse brain, injecting into the left side, left hemisphere. You can see the controlled mice we didn’t see any florescence. We choose a unique model here. This not only just efficient but very fast. About two hours it starts to work. Of course we need to demonstrate you can deliver this into the brain. We want spinal fluid delivery. We test this experiment on the mice. STEP-RNP was injected into lumbar puncture. You can see STEP-RNP penetrated the brain tissue. By hour 24 we cannot detect the RNP CRISPR in the body anymore. So this could be good. As I mentioned, this is bacteria. We don’t want it to stay in the body for a long time so this could be ideal. The question is whether this short exposure is good enough for treatment of Angelman Syndrome. Over the past two years in my lab and also with the team STEP-RNP lab tested this STEP-RNP in mice which mimic in human patients. We’ll tell you the findings in animal models.
>> Thanks. While I was sitting there I realized I’m the only biomedical engineer in this community. Any others can raise their hand. Welcome bioengineering to this amazing community. He’s the inventor. I’m the user. That’s how the relationship goes because we would not be able to stand here doing this amazing work. He has many patent. I have zero patent. I used to take five or ten minutes to describe this slide many years ago but now many of you, almost every one of you are very familiar with this concept. The idea is to use ASO or drug to reactivate the gene from chromosome. The beauty of the CRISPR is you may only need one dose, one spinal tap, permanent. That’s our hope. The idea the use of CRISPR [ indiscernible ] and reactivate the gene contribute to therapeutic benefit. We indeed are using human neurons. We put this new technology oftentimes to make sure, again, after we talked people came to me and said this is a nano particle. I make clear, it’s non viral, not nano particle. That’s why we give the name STEP. Clearly it’s very efficient. I misspoke on Thursday. The timing for this technology basically this is human neuron we generated and you can easily see that color. You can see almost no expression in the neuron. That shows how efficient this technology is. Also we show nicely in vivo, as Allyson said we do need mouse models sometimes for different experiment. This one shows you even the 90 day after one one, day two injection the spinal tap equivalent in human which is inject one dose of STEP RNP you can see clearly we saw the injection here at zero, no detectable. Injection roughly you have about 30% to 60% react evaluation so that’s pretty efficient based on all we learned from what Allyson talked about yesterday. A more high power view you can see this efficiency. It’s amazing efficient whenever we show this slide to our colleagues, people always amazed how efficient this single dose of injection, about 75% of neurons were edited by CRISPR. 76% of neurons we have react evaluation of the UBE3A gene. Of course you want to see whether this re activation can rescue or correct the phenotype which is a concern for many patients. This is a complicated slide prepared by the research associate. She’s here. She gave a talk yesterday. Last December we presented, this one year of work, a lot of work summarizes this slide. I won’t be able to go detail for each experiment the results but show a few example. You can see an open theater and put the mice in open theater and look at motion activity. Here’s the wild type normal mice. Here is the mice with Angelman Syndrome. You can see the reduce of motion and day after injection you can see that pattern is visible and tell the correction for this particular behavior. I’m going to skip this one. I’m going to show you a video. Allyson showed you a video yesterday. This is a day one day two injection. I think it would be better for me to show you you saw this one yesterday. This is the one I want to show sorry. Can we play this video? This is day 42 treatment, still correct a lot of our function behavior as efficient as day one for this particular behavior. That’s kind of important but of course this is in mice. Mice is mice and human is human. Mice is not human so make sure that’s interpreted the correct way. At least it shows the mice that treatment can correct behavior phenotype. As a research community we correct hundreds of mouse models but we also struggle in human. That’s the difference there. We cannot play this video sounds like. We also look for the intellectual ability and use multiple testing. This is a complicated slide with multiple replication. It’s a beautiful replication and we have very high confidence that this works efficiently like different time points, day one, day 21, day 42 and the paradigm. This is novel object recognition you heard yesterday. It goes through the correction of short term memory versus long term. In short term memory it’s effectively corrected, long term memory not too much. That tells you that correction somewhere is domain behavior specific which is not a surprise. The last one is the seizure. Of course this is one of the important feature, oftentimes challenging feature for many of the family here. Treatment for other medication may not be effective so treatment for STEP RNP gene correction can suppress all seizure phenotype pretty efficient based on the treatment at different dose and different timing. I think I’m going to pass to Allyson. >> This is meant to show the next step which is how do we get from all that incredible work that showed proof of concept that in both a newborn, a young and adult mouse model we can rescue almost all of the phenotypes of the behavior of the model which gives a lot of proof of concept that we can actually take this and bring this to patients once we have a human candidate so the human sequence in which we can employ this type of therapy. We’re at that point now and what the next steps are is how do we get that from the mouse to the human. Those steps is what we term translational research. I want to give a quick description of that and have you understand why it’s so complicated and also why it’s so expensive. These steps are really what it takes to get an investigational human drug or by biologic that is declared from work in the animal model to humans through the regulatory authorities. Experiments that are meant to evaluate for critical information related to toxicity, target engagement, off target risks and overall safety for the human application. These studies are received to as IND enabling studies and we talk about that a lot but let me share what that entails. Ultimately it involves many activities that are rigorous in order to get to that point and we do them often in parallel but by doing them in parallel in order to improve your timelines, it winds up costing a lot of money. So it is one of the most expensive parts of the drug development program prior to the clinical trial. There’s different pieces to this. The first is, is the drug working as you anticipated it would. The next is biodistribution. Where does the drug go. Where is this candidate in the brain in order to theoretically provide the benefit that you’re hoping for. Can it turn on the paternal allele in all of the vital brain structures. Then toxicology which is, is the dose levels you’re testing safe and that will drive the dose levels you can translate to in the clinic to ensure you have a nice margin of safety to say you’re not going to go into the clinic with something that could have incredible risk. What are all the things that are needed to prove that it’s safe. Finally, are there toxicities called Gino toxicities to the genome, the chromosomes, other genes around it. Will it change the genome in a way that’s irreversible that we have to be concerned about. These are all the studies we have to do in order to get to a human trial. So I want you to see how complicated this is. This is why things take a long time even though we’re all equally impatient. This is an example of a drug development plan and all of the work that needs to be done to get there. This is just to show you these arrows, one thing relates to another relates to another. We have to do that before we do that. We have to do this in order to do this. The only way we get to first in human is to be able to have all of these parallel pathways meet. Each one of these pathways has to have a go decision. There’s a chance that each one of these steps, that there’s going to be something that we need to pivot from, that doesn’t make sense. That is not going to allow us to take this path forward. So this next huge step is really vital to understand if we have a path forward to humans. Over all studies that need to bring our human candidate to human application takes a lot of coordination, a lot of time and a lot of money. We’re obviously very careful as we coordinate these types of studies because they’re critical but we want to do them in parallel in the right way. That includes the toxicology as well as the manufacturing, the regulatory component of engaging the regularities, and then the clinical trial pieces that we ensure we’re running the right type of trial. During every step there are going to be unexpected findings. That’s why it’s important to have a very nimble team that can work through those findings and ensure that we understand how to navigate them. We’re excited because we have a grant that’s allowing us to take this work and leap forward from mouse to human and the NIH is providing a large portion of the cost in order for us to be able to do this. There’s going to be a lot more that needs to be done in this drug development process that they’re not going to pay for that we’re going to have to support if the decision is made to go forward. Let’s have Liz talk about how we’re going to bring that information to a clinical trial. >> If these are parts of the Uber, I’m the destination. This grant is set up so you get all these preclinical pieces done and then we run a first in man clinical trial which is going to be a pretty intensive kind of thing. I’m going to talk about what that might look like. The grant required that we study two diseases or we couldn’t get funded so we are studying two diseases and we’re going to parallel through this process. We don’t actually start until about year three of the grant when we start planning and assuming that those other pieces go okay, then we can bring this to people. We need approval for the FDA, protocol approved, our IRB approval. Then we’ll no doubt, because this is a new compound, will be enrolling adults first for safety reasons. We’re going to try to run the trial at Rush and Yale and we can train some of the Yale people in anything needed to do this trial on Angelman Syndrome. We know this is going to be interthecal deliver. The location is probably based on preclinical data. We could use this by infusion but it could be ICM as you heard about in the back of the neck. This is one and done so you want your best delivery system for that at that time. We expect a long term correction. This is what the trial design will look like and this is a standard design for things that are very new and in phase one. We’ll be dosing one patient at the beginning and running these two identical trials and then have one patient and there will be a safety period. If that patient is safe we can enroll three patients. This will start with low dose. If these patients are all safe we can go to the middle dose and the high dose. If a patient has a dose limiting toxicity or a bad side effect, we can expand and do more patients at the lower dose. If two patients, say, at this dose had toxicity, we might have to go back a dose or stay at that dose. We find what dose might cause problems. After we get adults done we can move to children and do the same process with the children where we enroll one patient and then more patients in a systematic way to make sure this is safe. It was fairly easy to write a design for what we want to do in this trial because we have other trials going on in Angelman Syndrome. So we’re going to largely use the same measures that we’re using in the ASO trials and we’re going to have a lot of safety and tolerability tests that we have to monitor, much the same as all the ASO trials are doing. It’s important to point out that many of these measures were developed from ABOM data and natural history data and there’s the disease concept. This work is done ahead of time which made it easier for us to get this grant. We have to study H14 but fortunately that disease has a level of cognitive function similar to fragile X which some people know I’ve been working on for 20 years. We’ve done some validation and adapted certain tests like the cognitive battery and expressive language sampling for fragile X and we were able to make the case to roll those over and use them for this other form of intellectual disability. This work where we adapt appropriate outcomes and trial designs and accelerate new therapies to patients by getting early trials done is really the mission of our Rush pediatric neuroscience and this new grant capitalizes on the fact that we have the infrastructure to do this kind of thing. Hopefully there will be a next phase of development in this novel therapy. This is a perfect of example of how supporting the infrastructure and collaboration allow us to make scientific progress, brings teams together with various expertise to allow efficiencies that couldn’t be achieved in any other way. The work we’ve done in the past to create what we needed allowed us to get this grant. FAST works hard to ensure that we’re the low hanging fruit to get funding from either industry or NIH to look at these proof of concept early trials and have the clinical capability to move it forward. Thank you to everyone, NIH, FAST, Yale. This is going to be a really important collaboration.
