 Good afternoon, everybody. So I want to thank folks for joining as well as our guest, Dr. Sandesh Nagamani, and we'll introduce him at greater length in a minute. But I wanted to talk to you guys a little bit as an introduction to what we're calling the Excellence in Clinical Research Seminars. And although there's several names on this list that I recognize and have worked with, several of you I haven't, but I'm Dr. Chris Hours. I'm a pediatric oncologist that works with less besicker in segmental overgrowth disorders and investigating therapeutic trials in those disorders. And so this has been a particular topic of interest of myself and I think in the broader genetics community. Next slide. So Oleg and I were talking about, you know, ways that we can enhance clinical research education. And I think these seminars are one of the ways that we've thought about doing this. So our goal here is to recognize that genetics is not only a diagnostic subspecialty, but also now a therapeutic one too, especially with many of these new technologies. And, you know, these are sort of the motivation for this talk and future ones to get trainees exposed and also among faculty within NHGRIs intramural program and to talk about these topics more. So the, you know, the examples I have on the slide here, some of you may know, are recently FDA-improved medications or gene therapies that have arose from this, you know, application of genetic knowledge. And, you know, we say clinical research, it's a very broad umbrella. But, you know, I think these seminars, at least at first, are going to focus on interventional clinical research to sort of focus us and more specifically in clinical trials. But we do hope that this is an interactive and, you know, worthwhile thing for folks to participate in. So please make use of the chat box. So Oleg and I will monitor that as we go along. Certainly you can unmute yourself to engage in conversation, which is what we hope this to be. And a few suggestions for future seminars or topics of interest or just want to chat with one of us about clinical research. Please do the next slide. So just very broadly, some of the topics that we consider to be in this umbrella of things that we're talking about, you know, even clinical trials itself is a pretty broad thing. And so I've listed some things here such as study designs or even preclinical research and that first act of translating bench discoveries into the clinic. And then there's other, you know, important aspects of not only the science, but the involvement of regulatory agencies like the FDA. How do you manage all this clinical data? There's, of course, the issues of funding and budgets. The fact that some of these can't be done without industry partnership. And, you know, also, you know, in this understanding that this is a human subjects research endeavor and ethics plays a major role in what we do. Next slide. So for today, we played a focus on the basic question of how a study is built from its initial research question. Talk a little bit about the team required to design a study, the overall trial designs such as randomization with placebo control. And then there's others out there other than the gold standard randomized controlled trial. A little bit about eligibility criteria. You know, when you build a clinical trial, what disease are you studying and what types of patients are you studying? We'll touch a little bit on outcomes and endpoints. You know, those are our fancy ways of saying, what are we going to measure to know that our intervention is successful or not? Or, you know, having the intended consequence? And then, of course, from that is the hypothesis of what will change. And, of course, there's many other things that we may get to today. Or, you know, I think each of these bullets could be a whole hour discussion in itself. And so we'll frame this by starting with the journal article that was sent out previously, but much more in a conversational manner of one of the folks that led that effort and that study that we sent out. Ulick, I'll pass this to you now. Okay, well, thank you so much, Chris, for this wonderful introduction and it's my great pleasure to introduce and to welcome our colleague and our long-term friend, Dr. Sandesh Nagamani. Dr. Sandesh Nagamani is an associate professor at the Baylor College of Medicine in the Department of Molecular and Human Genetics. He is also serves as vice chair of clinical research in the department. Sandesh has a lot of experience with clinical trials and it is my great pleasure to really cherish this opportunity to learn from Sandesh and his experiences that he had accumulated through his involvement in many clinical trials and many aspects of regulatory medicine. Today, we'll focus on unpacking a lot of content that is hidden in the paper that was published in JCI in 2014 on describing clinical trials, a clinical trial of teraparitide in adults with osteogenesis imperfecta. The reason why we chose this paper is the following. It is a seminal paper, as of that day, it was the largest placebo control randomized trial in adults with osteogenesis imperfecta. It was, to my knowledge, the first trial in OI involving an anabolic agent. So it was a novel agent although it had some interesting history that supported its potential benefit in this patient population. The study was sufficiently powered to detect difference in the primary outcome. But of course, the question that I had and I will save this big question for Sandesh until a little bit later, the big question is, was it the data that you had accumulated? Was that enough to progress from IND, the initial new drug application? I'm sorry, from the IND to the, to NDA, new drug application in order to be approved by FDA for the indication for which he was studied. I'm gonna pause here and I really wanted to create an opportunity for Sandesh, maybe to share a couple of things about himself. And most importantly, Sandesh, I have a big question for you. How did you get interested in clinical trials in bone diseases? Again, many thanks to Chris and Oleg for this opportunity to talk to all of you. And what I also wanna say is Oleg and I were co-fellows together, so we go a long way back. One of the reasons as to why I got interested in rare bone disorders is that Baylor was a big center for osteogenesis and perfecta and our primary mentor that Oleg and I share had sort of two major interests. One was the inborn errors in metabolism, uterine cycle disorders and I do a lot of clinical research in UCDs as well and the second one was rare bone disorders. And so I chose based on the specialty that was available but also what really fascinated me was, I was always interested in skeletal biology and skeletal medicine, so understanding the rare bone disease as a model to understand more common disorders such as osteoporosis was really encouraging and that's why I went to do on this. Okay, thank you, Sandesh. I wanted to maybe set the scene a little bit and to refresh my understanding of osteogenesis and perfecta. So if you don't mind, I would like to spend maybe that's a couple of minutes reviewing this disorder with you. So osteogenesis and perfecta is a heritable form of connective tissue disorder. It is relatively common, one in 15,000 live births and it has variable presentation. The four hallmarks of these disorders is fractures, dentinogenesis and perfecta, hearing loss and blues clara. Now, variable presentation and it's probably pertinent not only within each type of osteogenesis and perfecta where you see variable presentation but you can also see quite a different range of manifestations from one type of OI to another. But it sounds like when you look at the variable presentation, what it looks to me, Sandesh, is for you focused on one particular outcome that we see in osteogenesis and perfecta and that is fractures. Is there a particular reason why it was or fractures or bone health? Is there a particular reason why that was selected as one of the main outcomes to focus on? So I think it's because of two reasons. One, in spite of OI being heritable connective tissue disorder that affects many of these and others such as muscle weakness, ligament to salacity, cardiopulmonary outcomes, skeletal deformities are the number one cause of morbidity in this disorder. Recurrent fractures are a major issue. And second, in this trial, what we were doing is to look at repurposing of a medication that was used to alter bone remodeling. It was not a molecular mechanism specific therapy. So it is very unlikely that that therapy would affect either hearing loss or cardiopulmonary outcomes. So those were the reasons why we limited it to the bone. Thank you. And one thing I realized is that when I trained and that was not that long ago, we really had, at that time, only three genes to interrogate if we suspected clinical diagnosis of osteogenesis imperfected, that's QOL-1A1, QOL-1A2, and CRTAP, DAP, at that time. But since then, a lot of things changed. And what I see right now is that this is the illustration that the buyer from one of your presentations is that I see there are quite a few disorders that are linked to the same phenotype. And does that make a difference in terms of selecting patients for the clinical trial in terms of the prevalence of the disease, the mechanism of the disease? What implications does it have for, and also maybe there's another aspect, is that how do you confirm that somebody has that diagnosis in order to be eligible for the trial? Right, so all I can say that this was, so these were slides from 2014 or 15. So there are now at least 18 plus genes known to cause the many forms of OI. Now they're counting 18 genotypes and more. But all it brings up an important point is to which patients do you enroll? In one way, over 90% of individuals with OI, they have either type one collagen structural mutations or mutations that affect the post-translational modification and processing of collagen. So 90 to 95% of all OI still occurs due to type one collagen-related mechanisms. The others are rare. Most of the trials that are ongoing are addressed the type one collagen-related OI, not only from the ease of enrollment, but also from the larger population they can target. But as we understand mechanisms-specific therapies, so there's an ongoing trial we're conducting now which a molecular mechanism therapy for TTF-based inhibition. We've limited enrollment only to genotypes-specific for triple helical domain of type one collagen or the CRTAP-403-Cheparone complex. Okay. So, and that leads me to this next slide. Again, that I borrowed from one of your presentation. There's about molecular mechanisms leading to the fracture and that really underwrites the rationale for choosing a teraparitide as an anabolic and chemical and also maybe we'll highlight this potential limitations with regards to this ability to improve patients' outcomes. So it sounds like if we really look at the mechanisms of OI, it really emerges at the interface of three mechanisms, not necessarily sort of excluding each other. They could be mutually reinforcing. It's one of the abnormal structure of the fibrils, altered interaction with the matrix protein in the bone and connective tissue and change cell-cell and cell-matrix interactions, all of them resulting in the decreased bone mass, which leads to fractures. And I really found this particular slide. Again, I borrowed that from your presentation. One of your past presentations is really highlights the fact that the bone health and the strength and the bone mass is really exists in the balance between the bone resorption and bone formation. And there has been at the time when you initiated and started this trial, medications or drugs that could be used to decrease the bone resorption to by inhibiting the activity of osteoclasts. The bisphosphonates is a classic example of that. You know, it decreases the bone turnover, increases bone marrow density, decreases pain and increases quality of life. But they had not been any agents at that time, which would focus on increasing the bone formation through stimulation of osteoblasts. And they had been prior experience with pteroparitide as a metabolic agent to stimulate bone formation. Just to remind everyone that the pteroparitide refers to the first 34 amino acids of the interminous of perethyroid hormone. It has an interesting regulatory status. It has been approved for the treatment of osteoporosis in women after menopause. It has shown to increase lumbar spine bone marrow density, modus increase in the hip bone marrow density and dramatic decrease in the frequency of fractures. In fact, one of the studies that I showed that I found shows a 50% decrease in the fractures in the posthumous of women. But then of course that sort of, that provides the rationale for, and to see if whether this chemical can or you know, peptide can increase the bone marrow density in subjects with osteogenesis imperfect by stimulating bone formation. And this is where we get to this, you know, a critical point in devising a clinical trial. And I really would appreciate your insights. And Chris, I know you spend a lot of time thinking about this particular aspect, you know, the critical decisions about the regulatory pathway, study design, sample power calculations and cohort formation of the cohort to exclusion and inclusion criteria and really focus and really defining the outcomes bringing your secondary and exploratory. Chris, you want to, I'm sure you've, I know you've said quite a few questions to highlight some of this, some of the aspects of clinical trials. Yeah. And you know, I think you've well laid out in the past few slides sort of this rationale for choosing a particular agent and the target population to apply it to. I guess one of my first questions was once you have that question that and you know, a population and an agent, what were some of the other sort of initial steps that took this from an idea and research question to building a team that could approach this with a clinical trial. What was the, I guess the team building aspect upfront to who was involved with players and led it to fruition? So one additional thing about the current therapy for OI, the specialists have been very well studied in children with OI, but when you look at the adults who already, some of whom have already received years of therapy, they may not have, they may not respond the same way. So we were a bunch of adults trained OI related docs who were, who got together. So this was the efforts were led by Eric Orwell at OHSU and Dr. Jay Shapiro at Kennedy Cree here and had a longstanding history in adult OI management of patients. And our center again had cared for a lot of patients here with adults with OI. And so the idea was to look at three large adult centers that would have enough patient population to enroll, who also had expertise in OI. So that's how we ended up coming up with this. What also helped was that these centers were also involved in what is called a linked clinical research center for osteogenesis imperfect is something that was driven by the patient advocacy group. So we did know, we had participated in longitudinal sort of cohort studies. And since all of us worked with this patient advocacy groups as most of us do in rare disorders, we worked under their chair leading to make sure that we moved this forward for a anabolic sort of therapy in adults with OI. Yeah, I'm actually interested if you could speak a little bit more about the patient advocacy side of that, those initial steps and clearly these groups are vital for rare disease research. I'm curious to what extent were they involved in some of the trial designs or the intervention itself or how it was being applied or measured and what kind of input did you or the team building this trial take from advocacy groups? So for this particular trial, I say that much of the design was done by the investigators but what really we appreciate from the patient advocacy groups is A, is there actually a need for another sort of medication? And what particular outcomes are they most likely to find meaningful? And this would be important in developing endpoints and assessing the right sort of outcome measures. And also the feasibility of this, where if they have to come in these number of times, do you think we will have enough patient engagement? And also for the enrollment, would it be possible for the patient advocacy groups to sort of let their members know that there's an ongoing trial which would sort of help enrollment? Now, some of the patient population may be wary of participating in the trials just because they do not know what it entails or they may be concerned about the adverse events and having a patient advocacy group that looks out for them to say, listen, this is the trial that is being done within the umbrella where we know what the investigators are doing. I think that helps immensely in enrollment of patients with right disorders. And I just wanna maybe draw the attention of our trainees is that if you really carefully think about how many patient advocacy groups are out there, the not every disease or group of disorders actually has a patient advocacy group representing their interests. So, and I think a lot of very motivated physicians and healthcare professionals, nurses and nurse practitioners who we work with can play instrumental role in jump-starting some of those efforts. I wanted to really focus on the first slide that I borrowed from your paper. And that is the study design. I have to say that there are very few examples, I think, out there and especially in the realm of rare bone disorders where randomized double-blind placebo-controlled study design is used. And I know that having been trained in the same sort of educational space with Sundesh, I know it is obviously a favored study design although it is the hardest one to implement. Can you share your experience with organizing randomized double-blind placebo-controlled RCT trials in rare diseases? And if you were to fix one problem, what would that be? So, I don't know. So for the fix one problem, is it scientific or familiar? I can give you the latter answer easier, but for this, if there are very ultra-rare groups, you're absolutely right that it's very hard to do randomized double-blind placebo-controlled trials. And in ultra-rare disorders where you may have 10, 20, 30, 40 patients at the maximum, and again, because the response may not be equivalent and same magnitude across the spectrum, one of the enrichment designs wherein you sort of subset a cohort of patients who are more likely to design and then randomize them, that would be much better. An alternative for the trainees is that when you have a smaller group of individuals who use what is called a crossover design wherein each patient is assigned to both a placebo and an intervention, but they first get the placebo versus first get the treatment and they're crossed over to get the other treatment after a washout period. Why we couldn't do that here is that at least in teriparicite, in the adult population, once you give this for a span of one and a half years, there is a prolonged fractal, the bone density does remain at least stable or the effects are there for years. So the washout period you would have to have is quite a long period and that would be impractical. So that's why we have to do a paddle and groove design. Now we have to make sure that this was randomized because the outcome measures should not have been biased by what treatment they got into. But there are a lot more studies now that are coming along with the double-blind placebo control trials, albeit with the smaller sample sizes. And I guess there's another aspect for some conditions where natural history is relatively well understood, you can use historic controls. That's another way to maybe adapt your study design to the realities of their rare diseases. They have been some discussions in the past about using asymmetric assignment to placebo versus teriparic or to active agent to maximize the group of patients assigned into an active agent. What are your thoughts on it? I think that's a great idea, especially if you're comparing it to controls in order to increase the sample size, you can have two, three controls for every patient that you had or you randomize a lot more number of patients on the intervention versus those that got the placebo. But the problem is because you're having all these, on the bone density part, you also need to have precision, right? If you have too few in one versus the other, you may underestimate or overestimate. So that's why we wanted to make sure that at least there's equal representation on both of those groups. Yeah, so too few patients will result in wider variability, your standard deviations will be wider and that in a while you think you create a tighter numbers for the intervention, wider standard deviations in the placebo control group will sort of even everything out. So it only works for some degree. And I think you're making an excellent point that these decisions of what to measure, how to measure it and what kind of randomized trial you're doing are all interconnected and inform one another. So a randomized controlled trial with a placebo is a decision not only based on scientific integrity but also what you're measuring and whether that's applicable. I think it's important for trainees to know that the RCT is what you hear about it most commonly but there are many different ways of skinning this cat including the late starts with draws. You can randomize not only the patients but timing. Sundesh, you mentioned the crossover study and I list these just to say that there's many different paths that people are taking but in the context of what you're measuring in the target population and ensuring that it's applicable but there's a lot of variety there to explore. Now there's a question from Dr. Solomon, clinical director of NHGRI. Hey, thanks and thanks for coming. This is great and a terrific opportunity for everybody including myself. I wanted to follow up on some of the questions that Oleg was asking about recruitment, Sundesh, if that's okay. I see all the time and Chris is probably more familiar than I with this the data that for example, for cancer trials everybody's taking part of them if they're fortunate enough to live near an academic medical center if they're by the coast and so on and so forth. I was curious about your experience with this trial or more genuinely in trials with OI or with other things. So to what extent are folks lucky enough to be around Baylor or Hopkins or OHSU or there's other things or traveling from all over and I know it depends it's very different Chris with proteas syndrome but I'm just curious to see big biases where people are from and so on and so forth with these conditions. Maybe just to echo Dr. Solomon's comment is I'm looking at the dropout rate here, about 20% this extra commute. I'm sure there are trade-offs. You're right. And again, I think that if you were to look at some of the ultra rare disorders that we do many of them, I would say probably 70% of them are from outside of Houston and they travel either by air or by road. For this particular trial at our side, I would say probably if we consider Texas, Louisiana, Oklahoma as one big catchment area and one single local area, we may have had two or three stations locally but majority of them came from elsewhere including the Midwest. And one thing that I find out is because of the logistics of getting some of these therapies especially when you don't have a regular job and you may not be able to do that regular job, many of the patients see this as an opportunity to get a treatment and so they are quite motivated and they travel long distances. So many of them did. Now, I think Chris had asked about is it ethical that you put them onto a placebo group? Shouldn't they be on some sort of treatment? Again, there's no standard of care in adults. And if you were to look, most of these patients were not on any medicines before but what we did was those patients who were on placebo after the trial ended, they were given an opportunity to go on the drug for the next 18 months on a per year basis. And also I put in the chat a link to a alternative design for clinical trials and rare disorders that will show you some of the enrichment sort of designs which the fellows may find interesting so that's a review article that I just sent. Oh, American Journal of Medical Genetics. It is, it is. Here you go. Dr. Solomon. Well, deep within the methods section, Sandesh, I believe is hidden what the most critical piece of study design that I think unfortunately is often overlooked by many of us, including myself. But I think if we spend just maybe half an hour on power calculations, it really changes the trajectory and allows us to maximize the use of clinical resources and patient population in a meaningful way. Some of the assumptions are pretty fundamental and then I don't think they changed too much about the alpha, 80% power, 5% difference. That's sort of a big question, right? How do you define clinical difference versus biological difference? And I think that's outside the scope of the study but then I really wanted to maybe focus on the 15% subject dropout rate. Because we talked a little bit about this. For example, for gene therapy trials where there's only one injection involved, the dropout rate is probably gonna be very, very small if anything as opposed to patients who have to receive daily injections plus travel to the clinical center. I noticed that in your calculations, the subject dropout rate was more than 15%. When you reflect back on the study design, did you feel that it could have changed anything if you were a little more generous and maybe 20, 25% to subject dropout rate or in the end it felt like it didn't really make that big of a difference? That's a great question, I'll take it in two sort of different ones. One is like, what did we calculate the power for? We say a 5% difference between treatment and placebo. That's a good number but what does that mean for the patients? And that's an important one for power calculations. Now, so let me actually preface by saying we don't know what a 5% increase in bone mineral density does to the fracture rate in individuals with OIs. In the patients who are from the osteoporosis world and the older one, a standard deviation decrease in the bone density, which equates to around 10%, that is associated with a 50% increase in the relative risk for fracture. And so it's unlikely that you're gonna get a 10% increase on bone density and even with the most robust agents you get somewhere between seven and eight excepting the newer ones. But it's possible that typically you can target for 5% bone changes. And so based on the prior clinical experience and pre-clinical studies, 5% change in the bone mineral density can actually translate into meaningful clinical results. In the non-OI population. Non-OI. In those with metabolic bone disease and osteoporosis. We're trying to look as to what that would mean in the OI population. So we have large natural history cohorts we've looked at. And in children, not adults, if you're less than 14 years, one gram per centimeter square increase in bone density reduces your relative risk of fracture by, I'm sorry, I misspoke, not relative risk. It reduces the association of fracture by 25%. It's the marginal effects on the regression. But here we had to pick something that would be achievable as well as something that at least in the adult population has shown to reduce fracture risk. So that's why 5% was chosen. The second aspect, all I've heard about is that, yeah, you said 15% subject crop out, right? But it was higher. So were you powered enough to do this? And he's right in the sense that at the end when we analyzed, we analyzed less than the number of individuals we did, this again, when we did an interim analysis, we did see that there was a realistic probability that we would achieve that number even with fewer patients than 19. So at the end, it did not make a difference. However, these are considerations you have to put in while calculating the sample size. Did you not see a difference because you were not powered enough to do it or was there no actual difference and that was an important question. Yeah. And then there's another thing that constraints further may constrain the pool of eligible patients. And I have one question. I'm sure that Chris has another question for you about this. And that is, I noticed in the inclusion criteria, you emphasize that adults with a clinical diagnosis, oh, I think that's understandable. Why patients had to have fused epithesis? So one is that that's sort of a sign of pediatric skeleton versus adult skeleton. The second aspect is because pre-clinical studies with teraparitide and rat models have shown a high risk for osteosarcoma. And as you would know, most of these are at the upper end of the growth plate in the longer bones. So we wanted to make sure that we minimize sort of that risk in anybody who sort of has unused epithesis. So we, from a safety perspective as well as from making sure that they were adult skeletons without growing growth plates, we ended up having fused. So exposure to a teraparitide in individuals with active growth plates may increase their risk of what kind of complications? No, no, no. I'm not saying that they increase the risk of osteosarcoma, but osteosarcoma in the pre-clinical model, rat models. In rat models, yeah. Right, so when rats were treated between three and 60 times, the dose that is used in humans for prolonged period of time, they develop many bone tumors. So what we have to do when treating adults with osteoporosis also is to make sure that those with high risk for osteosarcoma, pageants disease, all those people are excluded from treatment. And here, because in a children, there's a bimodal peak for osteosarcoma, right? In the teens with the open up of the season. So that's why we wanted to make sure that we, from a safety perspective as well, we don't include them. I see. Go ahead, Chris. And I think from a safety perspective and also this question about osteosarcoma in a pre-clinical model, one potential advantage you had when repurposing an agent versus a first inhuman or earlier or unapproved agent was an existing safety data set and it experienced in the osteoporosis world. Again, non-OI, so maybe it doesn't translate perfectly but there's reason to feel that it does. And I guess my question is how does that influence your exclusion criteria when you're selecting who should not be on this agent and how was that informed? Let me unpack that for a second. Oh, like do you think you can go back one slide if you don't mind, please? Yeah, there you go. So Chris actually hit it really on the, on the head of the nail where it is much, it is a lot less generous if you have approved therapies because within the FDA, you have a drug master file all of the pre-clinical safety data as well as the human clinical data. So if you're not using doses that are far beyond what would be accepted normal from a safety perspective, the FDA would not have a significant number of questions and that I think really helps repurpose. Now, if it's a first inhuman, then you have to end up making sure that you work with the FDA to not only to first assess the safety in the phase one and two trials and I know Chris hates these, especially in long in rare disorder trials because there's a merging of all of these numbers but let's say safety trials versus efficacy trials. So for us, so we took a two prong approach. One, we excluded those individuals whom we thought would have characteristics that would interfere with the endpoint assessment. So if they've had very recent bisphosphonate therapy that could affect the bone number density. If they have spinal instrumentation that prevents us from measuring their bone density, we excluded those individuals. The second from the safety aspect is again, we went through to make sure that from the FDA prescribing information, you can't use this if people have abnormal alkaline phosphatase because that could be a harbinger of Padgett's disease. So we made sure that those people were excluded, those with abnormal liver functions and others that could contribute to the safety, we excluded them. But other than the good clinical practice of exclusion of those individuals with characteristics that may put them at increased risk, we didn't have to meet a higher bar for safety here than what you would encounter in the first human. Yeah, and before I launch into the results and I really had to select a few of them knowing that we won't have a lot of time to go over just a lot of data that's stored in both the paper itself, supplemental materials. But I just wanted to make a brief stop here at the trial endpoints. When I look at the primary endpoint and the secondary endpoint, I'm sensing that FDA would probably want to see self-reported fractures or the number of fractures in a given period of time as their primary endpoint and the change in the bone marrow density as measured by DEXA or QCT would be secondary endpoints. What was the rationale for choosing primary endpoint as they were? So you're exactly right. If the FDA were to review this and say, listen, if you're going to need a new indication for this you should show a fracture because they would have liked fractures as the primary endpoint. With 40 patients in each randomized groups you will not have enough events in order to pick up a decrease in fracture risk within a span of one and a half years. So at that time, we didn't have these numbers as to what would be the per person of fracture incidents and stuff like that. We have some data now from the European cohorts but that's one of the limited sample size. I don't think we would have been able to detect the meaningful difference in fractures which we did not in the cohort when we looked at the secondary endpoints. So that was the rationale to use a surrogate endpoint that could show a change for this approach. Oleg, I think you're muted. They're still muted. I think that's a very important thing that you just mentioned there. And I also want to make sure that the trainees understand too that when you run a clinical trial it's not always with a goal of changing the practice of medicine but it may be to figure out where to put further investment to that ultimate what the FDA may refer to as a pivotal trial or one that holds its own to say this drug should be used for this agent and also ultimately allows a pharmaceutical company to market the drug for that indication or in that group of people. And so Sundesh on the spectrum of completely exploratory to ready to present to the FDA for marketing approval where would you say that the goals of this trial were? I think they were somewhere in between of saying that because this approach was not tried. And the reason from a genetic standpoint, right? Is that those with more severe OI tend to have sort of a dominant negative effect. So if you make more abnormal type one collagen within the bone matrix would that actually translate to a better bone density versus not, right? So the idea was to look at what anabolic therapy does from a safety perspective, we were not worried from an efficacy perspective we needed to know this approach would work. And if it did work quite well the question was whether this could have gone in for an approval pivotal trial as you put it. The problem that the issue then would be that we would have had to discuss with the FDA what an acceptable end point is. And to this day in the rare bone disorders field we're frustrated by the lack of a surrogate endpoint that the FDA would accept for drug approval. They still want fractures which would mean a lot more number of patients. So I would say that somewhere in between exploratory and approval was where this was. And maybe just to make one quick comment about the surrogate end points is the discussion involving FDA about surrogate end points shouldn't wait until you get to phase three. Surrogate end points especially for rare diseases the discussion should probably happen at the pre-clinical stage before you get to IND because there's just so many things as you can see in this study primary and secondary end points and the ability to advance to the NDA stage. It depends so much on what FDA considers the big three outcomes which is better feeling, better function and proof survival. But if you can't demonstrate those three things which in this case would have been probably a decrease in the frequency of fractures but FDA probably would have to agree to use a change or improvement in the bone marrow density as an acceptable surrogate endpoint. But that's not something that needs to be saved until phase three or just about you submit an NDA. I think so as we I'm looking at these results I really feel like it's almost the fulfillment of the power calculations. You predicted you would be able to detect 5% difference with the cohort that you had and to demonstrate statistical significance that as expected by you. But maybe I wanted to highlight a slightly different so it looks like you're meeting your primary outcome for this clinical trial but I have a slightly different question. You calculated that you needed 19 individuals. You enrolled in a 78 individuals and I think there was an interim assessment. Can you tell us a little more about the circumstances under which you would say that an interim assessment in a control study would be necessary? So I think now for most of these larger studies interim analyses are being implemented either to understand whether it's going in the right direction to understand whether there are a group of individuals who respond much better so that then they can be put into an adaptive design. And so it's become a part of bettering our clinical trial sort of design. But what I would stress to the trainees is that if you do an interim analysis you have to account for the fact that you've already done that and you have to adjust that with an adjusted significance ratio. Is it because you unblind yourself? No, no, we do not unblind ourselves. So the analysis is done by someone who is now. It is an independent group. And what they will suggest is whether there is a reasonable degree of whether you are more likely to achieve it with a smaller number, more likely to achieve it with a bigger number or you've already achieved it. So we can, but again, as Oleg did say that we have to have this for planned interim analysis of specific enrollment points. We can't end up putting this after the trial design. Chris, do you have any questions about the slide before I advance to the next one? Yeah, no, I think the only thing I wanted to add to the question about the interim analysis is I think sometimes the way I think about it that may be familiar to geneticists is GWAS. You're doing lots of different tests at different genetic loci. And that's why your P value has to shift. And 0.05 is an acceptable in some of those analyses. Similarly, if you're looking at a primary endpoint multiple times, your type one error rates can change and this is part of the statistical underpinning that hopefully you have a bright aware statistician that's helping with these aspects of the trial and analysis. Okay, well, the next slide really reminds me of a very simple fact is that it's something we just discussed maybe 15 minutes ago is the change in the bone marrow density doesn't necessarily translate an increased bone strength. And I think, you know, Sandesh, you and I and Chris, we talked a little bit about this during preview of this lecture is that it's actually the first one that's not easy to measure in bone strength noninvasively. How did you achieve that? And do you think this method has the potential to be accepted by FDA in the future as surrogate endpoint? So you're right, there's no way you can do this in humans in the mice and the preclinical studies. Once you give the treatment, what you do is you basically do what is called a three-point benefit. You look at the bone here and you fix it in two points and then you pressure and then you see it, what load it gives way. So in humans, what you end up doing is you do a a corner of a computer tomogram and you sort of, you break down that 3D figure into rigid cubes or small cubes or small cubes and look at from a computational standpoint, given that this is what the structure is, including the travecular bone inside, at what load of pressure is it going to give way or deform? And so it's an estimated sort of bone, it's a calculated number using finite element analysis. And what has been shown in the adult population is when you do these FDA in a parametric model and you actually then do the three-point bending sort of pressure, there's a very good correlation between those and that's why you do this. Now the second question that I get asked is, is it possible to do the FDA accepting it? Now there's a trial with another anabolic plus anti-resorting agent put together, it's called a citrusumab in Hawaii, where they have looked at the estimated vertebral strength as well as volumetric bone density at the peripheral bone to see that at the radius or the tibia. And so those sort of discussions are now being had with the FDA. The problem is that you don't have these analyses that are available everywhere. In the United States, there are only five high-resolution QCDs that can measure at the level they want, which would mean that most patients would be excluded unless they travel to the sites. So right now I don't think the FDA has accepted that as a surrogate one, but the trial with citrusumab has used one of these estimated vertebral strengths as one of their primary endpoints. We got a question from the audience on, why do the total hip bone marrow density decrease in the placebo group after 18 months? As opposed to actually looking at the lumbar spine bone marrow density, which tends to appear to go up. Right. So lumbar spine is a lot more responsive because it has a lot more of the cancellous or lamellar bone, while the hip has both cancellous as well as cortical sort of bone. Now, I do think that there is always an element of variability that comes in with the assessment. And you would assume that that sort of should normalize over both the control and the population sort of groups. And or there is actually a decrease without any therapy because we all lose bone density after we hit 30 years of age because the formation is not the same as research. So the way to look at it, there's two things. One, that it could be a technical issue itself whether or not there is a problem. We're hoping that if it's a technical issue then all of them including the treatment arm should have been the same. The second, it could have just been a problem because of normal wear and tear of the bone. I see, thank you Sandesh for that. And then this is a very interesting slide that sort of really I had to pull some of these data from the supplemental materials. But really highlights if you subtract data obtained from subjects with OI type three and four and you only keep OI type one, you get even higher percent change from baseline. So what's going on here biologically? So there are some unpublished data that I think that we'll be able to share sometime soon. This may have to do with how the PTH signaling is interlinked with another dysregulated pathway within the OI field. But also I do think that one is type one OI, the individuals tend to have half blowing sufficiency and null mutations, right? So here they have half normal amount of collagen which is not dominant negative as opposed to the matrix is completely different in type three and four. So it may be completely a biological response in what the terepatternite does. I see, I see, well, that's fascinating. Just to be able to learn some of that biology through a clinical trial, that's fascinating. I have to say that after this trial, we've, meaning our group, Dr. Lee has led all of those studies here to look mechanistically as to why there may be a differential response in the severe OIs for the minor OIs using mouse models that recapitulate these forms of OI. Okay, thank you. Well, I'm getting to my last slide, but I think it's a loaded one. It appears that there were no difference in the observed frequency of adverse events or serious adverse events between the intervention and placebo groups. There were no serious adverse events, but there was one death in the placebo group. And I realized that the vast majority of rare diseases carry significant morbidity and mortality with them. I understand that it happened in the placebo group. What if it were to happen in the intervention group? Can you tell us a little more about the implications of this type of serious adverse events that happened during clinical trial? And some of you and I will live through some of those, right? I know, I know. I even know which one you're even talking about, but... Yeah, but can you unpack the implications of those things in the context of a clinical trial? I think it depends as to what specialty you're in. So in Chris's specialty, if you're a chemon sort of where you actually do see essays all the time, this would not be unheard of, but death in any clinical trial raises red flags. So what we have to do as investigators would be to see if there is a reasonable possibility that was related to the study intervention. So the relatedness of an S.A.E, be it a lab abnormality, a clinical abnormality, how was there a temporal relationship with the drug? And do we think it's from a potentially know what we know from the pre-clinical studies and the temporal association is it reasonable? If it is, then we have to report it to the, not only the local IRBs within an expedited manner, but also a data safety monitoring board that has to review the data. Now for this trial, the IRB at that time did not require a data safety monitoring board given that we were using an intervention that was FDA approved for similar bone related disorders and we were using the same dosage. But if it were to be a first in human drug trial, the DMC would have at least made sure to review this and give us guidance and what we need to do on progression of the trial. I see a question from the audience from tail. Do you expect the use of a peritide for two years or longer in patients with a Y, any concerns regarding osteosarcoma after prolonged use? Okay, so would you expect the use of terepidiotide for two years or longer? And the short answer to that is no. And the reason is the bone formation and resorption are always coupled, they go sort of hand in hand. So you typically most agents that increase the bone formation also increase the bone resorption. With use of terepidiotide, there's a gap or an anabolic window when the formation goes much higher than the resorption and you can build bone. But if you keep on giving terepidiotide for more than two years, that gap closes out and you're not going to get any additional gains. So that is why even in the osteoporosis world or here therapy is not for more than two years. Now that's not unique to terepidiotide, another anabolic agent called Romasusumab, an anti-sthorostin antibody has been approved for adults with osteoporosis. And even there, the therapy is limited to one year duration because the gap in which you can build bone would close after that. Now regarding concerns for osteosarcoma, again, this was in the preclinical models in humans this has not panned out and recent publications have shown that there's no more than a background sort of incidents of osteosarcoma and those who have used terepidiotide versus the general population. But again, as I said, if I do think that some, so for example, in individuals who've had bone irradiation for a whole beam x-ray irradiation, you can't use this medication. In those who had bone cancer cancers and which have metastasized to bone, I'm very hesitant to use an osteoblastic agent just because even if they're not in a cancer cell somewhere and you don't know it, I don't do that. So for those who have cancer who had a risk of osteosarcoma, have this disease, I make sure not to use them in practice. And I think we have just a very small amount of time left. So I was wondering, Sandesh, if you could tell us a little bit about the overall results of this study and how do they inform current or future investigations in OI and clinical trials? So the overall results actually does show that antibiotic therapy would be something you can try in OI, and it is more likely to affect the mild OI than in the more severe forms. While I don't think we can claim credit for it, there are other antibiotic therapies that have come and have been investigated now. Francie Zidlohieu, who many of you might know, he has been a pioneer in OI research and who introduced the persuasiveness therapy has led to the clinical trials of cetrusumab in OI, and they have finished that, they are now the largest OI trials, anabolic OI trials in the Bells. And I just saw an abstract from them which really ended up showing that anabolic therapy has the potential to increase bone marrow density robustly in both mild and severe OIs with cetrusumab. So I think that was one thing that we learned. The second thing we learned, and again, this may be important for the trainees, is that when I was trained and OI will let you know, we were asked to collect everything and anything possible and bank them after the experiments. And sometimes it is painful because we don't use that for most of the time, 80% of the time they've been made in the refrigerator where it takes a lot of time and that was a source of frustration. But for this trial, we had banked plasma from the trial and what are really, and these are efforts that Lindsay Nicholl and Eric Orwell have read that within the Brutal Bone Disorders Consortium is that they were able to go back and look at some of the biomarkers. We looked at what anabolic therapy does to overall other markers of collagen biomarkers. We were able to look at sclerostin levels and look as to whether this could be used as a biomarker in for the diagnosis or for treatment response with anabolic agents. We were able to develop some from the plasma biomarkers for growth in OI. So for the trainees as frustrating as it is to sort of collect more than what you would require, it's great to have all of these somewhere. And when you get an idea, you always have the samples to go back to do. So that I think was a major lesson for us. And again, since they're trainees, I have to say when I did this trial, I didn't have much knowledge about clinical research as well. I was a trainee in genetics just becoming a junior faculty. And what I took out of this was to learning from the senior investigators, Jay Shapiro and Eric Orwell and Brendan Gee, was to how actually to conduct the trial, what it requires and how I can further my career based not only on the insights I get from this, but to sort of leverage from it. So this trial gave me a sort of a platform to say I've done these sort of trials. And when I proposed these in my own grant applications, I've not been as, you, we've not seen any evidence of you doing it. So when you get opportunities to do things like this, especially at a place like NIH, I would really encourage you to take them with both hands. Okay. Well, Sandesh, thank you so much for joining us today. It was a great opportunity for us to learn from your experience. Thank you for sharing your experience. Thank you for taking us behind the scenes. I know that your paper really summarizes a lot of your experience, but there's a whole lot more that goes into clinical trials to make them successful. And we really appreciate the opportunity to take a deep dive into sort of that supplemental materials. Thank you so much. Chris, did you want to offer any concluding remarks? No, I can't add anything else other than more gratitude for joining us and sharing your experiences. It's a pleasure. And if I may want to put a plug in, so since we're not within the NIH, we're in the extramural world. So there are some of the trainee resources that we have developed with the generous support of the NIH. And one of them is called the OITELEECO-CD. So if you were to Google, osteogenesis imperfecta teleeco, and let me try to get the URL for you. So there is an OITELEECO-CD that basically has a lot of nice lectures and clinical management for rare bone disorders. So please make sure that if you want to look into that, there's also a rare bone echo, which talks about a lot of rare bone disorders. Many of the speakers are from the NIH. So it'll be instructional for you. So thank you very much for allowing me to spend some time with you. And just to follow up the Sunday session on teleeco, just posted a link to our Twitter where we featured the upcoming teleeco clinic series and rare bone disease, the dysmorphology exam for skeletal dysplagias by Dr. Danita Velasco. So she is at the University of Nebraska and the link will contain another link to for interested trainees to register for that event. Again, Sandesh, thank you so much for joining us. It was a great pleasure to learn from an experienced clinical investigator like you. And we're looking forward to hosting you in the future, knowing that you do clinical research, not only in bone disease, but also metabolic disorders. There you go. Thank you very much, guys. Thank you.