 Well, thank you. I too very much appreciate the opportunity to come and present. And as you'll see, pediatrics is pretty much my passion. The reason for this is that children with autism or epilepsy or asthma or inflammatory bowel disease or juvenile idiopathic arthritis, once they turn 18, they no longer become healthy adults. The consequences of administering nephrotoxic or ototoxic medications in the NICU do not magically go away when these survivors turn 18. And so my bias is that early and effective intervention may actually change the course of disease as these children become adults. I'd like to acknowledge the grants of support first because this generally is what I forget to mention. And interestingly, the bottom grant did not make its way into the list we saw yesterday and that's because it turns out, is it feedback? Can you hear feedback? A bit. Did not make its way into the report. And as it turns out, I didn't use the term pharmacogenetics or pharmacogenomics anywhere in the abstract of the title and that's because we're interested in precision therapeutics for children. And so in the time allotted to me, I'm really going to superficially address these four points and you can read them and they will come up in additional slides. But as I alluded to yesterday, in terms of implementation, we don't feel like we have much to implement until such time as we generate the information that needs to be implemented. And in constructing how it is that we are going to develop this information, we are sort of reversing the traditional dose exposure response paradigm to work backwards from response. And this is because we feel, the practitioners feel that when they implement a therapeutic innovation, whether it's small molecule or behavioral or anything else, generally they do so with some sort of outcome or response in mind. And so the questions that we work through are, you know, what is the therapeutic goal of the intervention? How much small molecule, active form of the small molecule needs to be in the system to have a high probability of that response? And then how does the dose have to be individualized to achieve that target exposure? And so just to move into this first phase, clinically useful guidance, here is information taken from the product monograph for atymoxidine. This is a medication used to treat attention deficit hyperactivity disorder in children. In fact, it was developed for this indication. And this is the verbiage in the product monograph. It says, poor metabolizers, PMs of CYP2D6 have a 10-fold higher AUC area under the curve, so a measure of systemic exposure and a 5-fold higher peak concentration compared to extensive metabolizers. Down at the bottom it says that there may be some dosage adjustment required. Actually, I ask pharmacy students when I present them this, okay, 10-fold difference in exposure, how much will the dose need to be adjusted in a poor metabolizer to achieve the same exposure? Apparently this is a very difficult calculation. And it appears to be difficult because when you actually look at the guidance, the starting dose is the same, 0.5 milligrams per kilo, whether you are a poor metabolizer or an extensive metabolizer, despite the 10-fold difference in systemic exposure. And we can go into why this may be, but from an implementation perspective in terms of helping practitioners guide the dosing of atomoxetine given information relevant to that individual, their CYP2D6 genotype, this is not helpful. And so we're approaching this problem by conducting genotype-stratified pharmacokinetic studies to try to get, using an efficient design, to capture how much variability there might be across the population. Now, when we compare, if you look at the median values here, there is a 14-fold difference between the PM population, or group, it's not a population, it's four individuals, in red versus the blues that have at least two functional alleles. Now, it turns out that it's possible to give kids 0.5 milligrams per kilo, but you have to mix and match the available solid or a dosage forms. So we thought, well, some of this variability may be due to the fact that the children didn't actually get exactly 0.5 milligrams per kilo. So if we correct for the dose that's administered, we can get that difference between PMs and EMs down to 11-fold. So that's pretty much what's in the product monograph. What's relevant from an individualized dosing perspective or a precision therapeutics perspective is the fact that this same weight-based dose gave a 50-fold range in exposures if you go with the uncorrected data, 50-fold, that's more than 1.5 orders of magnitude. But we can get that down to 30-fold if we correct for dose. Now, the way that community practitioners will prescribe this drug is what's shown there on the left-hand panel. So it's a 50-fold exposure for children given the same weight-based dose. So you can see where some of the challenges are now in terms of implementing a drug. As it turns out, and I don't have time to go into this story, the consequence of the FDA-approved dosing guidelines is that those individuals who are in green or blue are unlikely to achieve concentrations that have been associated with clinical response. So that's probably at least 40% of the population. And they are paying the price for making sure that those individuals in red do not have concentrations that are too high. So in the course of conducting some of these genotype stratified PK studies, we have stumbled upon some unknown unknowns and they have taught us an important lesson, and that is the limitations of using available adult data to inform the design of studies that we're going to conduct in children. And I'm going to give you one example of this. This is a study of Simvastatin in dyslipidemic children conducted by a member of our group, a pediatric cardiologist by the name of John Wagner. And basically, Simvastatin requires hydrolysis of an inactive lactone to the active acid. And just to cut to the quick here, we made the assumption that clearance in kids might actually be quicker than in adults because CIP3A activity is responsible for the elimination of the acid and the lactone. As it turns out, we were able to replicate the genotype phenotype associations seen in adults except the magnitude of effect between the TT homozygates and the CCC variant group at position 521 is six-fold as opposed to three-fold in adults. But the most important thing was that 25% of the participants in this genotype stratified PK study did not have Simvastatin acid concentrations that were detectable. And it looks like because the half-life, we couldn't get a half-life because it was basically a flat line in most of these kids. It looks like rate-limiting formation, so there is something going on in children related to the conversion of the lactone to the acid that needs to be looked into. But we would not have anticipated the possibility of rate-limiting lactone formation based on the available evidence from adults. So there is much more to learn. John is conducting, has completed a pravastatin study in these same kids. He's about 75% of the way through atorvastatin and bursuvastatin. And I can tell you that there is something new that we've learned with respect to pravastatin as well. And the question is, to what extent are these actually relevant to adults as well? You can find clues to what's going on if you go way, way, way back in the literature. We would have never looked there except for finding these results. Variability and drug response. This is another big bugaboo of mine in that when we are looking at drug response, there's at least two possibilities, and Dr. Weinschelbaum is out of the third. The first is that some people may not respond because of inadequate exposure. And I think the atomoxetine story is an example of this. Other individuals may not respond because there is something fundamentally different either because of genetic variation in the drug target or developmental differences in the drug target if we're talking about children. And we need to be able to distinguish between those. Dr. Weinschelbaum pointed out that administration of a drug can also add sort of unmask perhaps differences or be able to, we may be able to stratify a patient population by disease on the basis of response to a medication rather than by clinical symptoms. So anyway, being able to distinguish between the reason for lack of response is going to be very important because the treatment decisions will be quite different. For example, if it's low exposure, you can increase the dose. If somebody is inherently not going to respond to the medication, you'll want to change the medication, but ultimately you would like to be able to make those decisions before you prescribe a drug the first time. And so conceptually, this is a very simplistic approach to it. These are three dose response curves that are shifted twofold. And I picked twofold because the minus 16, 39 G to A polymorphism and morphine, if you go back to the original New England Journal of Medicine article, it was about a 1.8 fold shift in the level of messenger RNA expression for VKRC1 that was associated with that variant. And the point of this slide is simply to show you that if you target the same therapeutic goal, that the exposures that are going to be necessary to achieve that same therapeutic goal are going to differ depending on what the drug target genotype is or what the drug target level of expression is. And so this is one of the arguments for looking at drug response and looking at things there first and trying to control exposure to maximize the probability of clinical success given the drug target genotype. Finally, I think one of the challenges is considering individual patients, whether they are pediatric patients or adults as individuals. And here's an example. Here's what we usually do is we like to look at the distribution of patients who participate in our studies and we will do a frequency histogram. Here's one for age or one for height of 189 kids that participated in a CYP2D6 longitudinal phenotyping study that we conducted where we were looking to see if CYP2D6 and 3A activity changed as children went through puberty. When it comes to treating individual patients though, it's not so much where they are on a population distribution but what they look like as individuals. And so in this figure here, all I'm doing is illustrating the point that a child can be defined by a number of characteristics. Their height and weight had a particular age, whether they are male or female, so red or blue. And in terms of the size of the spheres, this is indicating development, stage of development, tanner stage, for example. Now if I could show you a sixth dimension, I could show you their CYP2D6 genotype or any other characteristic. But the idea is to start thinking of individuals as individuals and trying to characterize those factors that make them individuals if we truly intend to implement pediatric precision therapeutics. I have five minutes left. I can't believe I have five minutes left because usually I go over. So the take-home message is we are trying to develop this data set. We call the program Goldilocks. The concept is very easy to get across to patients and parents. Not too big, not too small. The dose of medication is just right for your child. It turns out the Department of Philanthropy can work with this quite well in terms of their engagement with the community to raise funds for the hospital. And so the Goldilocks acronym really is intended to incorporate those factors that make each child unique. Their genome, their stage of development, and to use that information to develop these dosing algorithms that ultimately will optimize drug therapy. As I mentioned, we are pursuing this response exposure dose paradigm. And that is we are focusing on the drug target genotype individualizing exposure so that we can minimize the contribution of variability in drug exposure as a contributing factor to variability in drug response. And this is particularly important for drugs like atomoxetine or any other compound that is subject to polymorphic genes involved in their disposition. Yesterday I mentioned one other thing, and that is that it's important to educate patients and their families about some of these rather complex concepts that we are trying to address. And so this is a colleague of mine, Sue Rahman, who has spent a lot of time working with the Art Institute in town in Kansas City and the students of that Art Institute to come up with ways of illustrating these concepts. And so here is the concept of genes in your body being responsible for breaking down a medication and that depending on what your genes are, you may be a turtle. And for the record, I am a turtle. I am a CYP2D6 poor metabolizer. You can be a bike. You can be a car. Or at the far end of the spectrum, you can be a rocket. And you won't be able to see it. But the visual, here we've got a person who's got a lot of drug in their system. They're not all that happy. And here is an individual that doesn't have very much drug in their body and actually they look frazzled. The people at the front will be able to see they're frazzled. This was tailored for our ADHD study. But basically, it is really important to be able to communicate, if we're going to generate the knowledge, we have to be able to communicate to patients and their parents how important it is to be active participants in this process. So just to close, I mentioned that we have embedded within our division pediatric subspecialists. They work together with individuals with expertise in pharmacogenetics, Andrea Gadek, for example, in vitro phenotyping. One of the other things that we're doing is looking to see for some drugs where it may not be possible to do a genotype stratified PK study. We're looking at creating virtual children using physiology-based pharmacokinetic platform built around drug biotransformation from individual pediatric liver and microsomal samples. This is actually working out not too badly for atymoxidine, analytical chemistry, the quantitative pharmacology. And then these are the individuals who are really involved with taking this Goldilocks approach into their patient population. You'll see that we have rheumatology, developmental and behavioral sciences, infectious diseases, allergy and immunology, neonatology, gastroenterology, adolescent medicine, oncology, cardiology, and rehabilitation medicine. That's a rehabilitation medicine is our newest one. But the idea, our institution is committed to moving this concept into all the pediatric subspecialties. We certainly appreciate their support and I appreciate your attention and hope you'll take home a passion for implementing pediatrics as part of your programs. Thank you. That has to be the first time I've been on time presentation. Great. Time for questions. Terry, yes. Steve, this is really exciting. I think you may have said this, but I missed it. You're currently implementing in these sites that you've, or these specialties, but your plan is to do it across all inpatient and outpatient? Yeah, so we're not implementing anywhere. These are the individuals who are collecting the data, collecting the data that can be used to implement. I would say that CYP2D6 and meds for developmental and behavioral sciences, so hopefully SSRIs and some atypical anesthetics, will probably, that and statins and dyslipidemic kids will probably get to an implementation phase sooner than some of the others. Let's see, the infectious diseases is actually a bacterium hypersensitivity focus, so that will be longer. So just as a follow-up, if you haven't implemented yet, could you consider randomizing so that we could actually generate some evidence from your experience? Yeah, I shouldn't say that we're not implementing. We do have an individualized pediatric therapeutics clinic where we do incorporate pharmacogenetics and genomics. I would consider that to be more like a forensic application. These are children who are coming with a specific problem and we get the pharmacogenomic testing done to help solve those problems. We also have a consult service. So it's more reactive right now. The focus is on developing the tools to allow us, we want to take this into the ADHD population first with atomoxetine. It's a third-line drug because our clinicians think it doesn't work. They would prefer to use it because that would allow them to avoid some of the behavioral changes that they see in kids on stimulants as well as the appetite suppressant effects. So we're not there yet, but we're moving there. So consider randomizing. Okay, our next question is with Bob. So pediatricians are used to adjusting dose, so it's not just one dose if it's all because they're different stages of development and different weight and body surface area, and they're used to doing that. Do you think that that makes it easier to get them to engage with this kind of dose adjustment for pharmacogenomics than for adults? Yeah, I don't know if I... In our place, I think it does for two reasons. Number one, our clinical pharmacology program has been there for 20 years, and we are able to do some of these studies, genotype stratified PK studies in kids who are not receiving the medications typically just because of the culture that's built up over those 20 years. So I don't know that everybody... what I'm going to say is generalizable to everybody. The other factor that I think will allow our clinicians to be more accepting of this is the fact that I'm not taking this to them. Those people are taking it to them. So it's a cardiologist who's going to help them understand SLC01B1 pharmacogenetics and statins, and CYP2C9 and VKORC1 in the Fontan patients who get warfarin. It's going to be the practitioners that do it. We've already found out that Jazzy Ann Tolbert, the oncologist, she's been... her colleagues have come up to her and said, well, we really like to know what it is that you know, but we don't have time to do like a two-year clinical pharmacology fellowship on top of the pediatric subspecialty fellowship. And so now we've Jazzy and another... excuse me, another member of the division, Jen Lowry, have developed a boot camp where practitioners will get an intensive course on clinical pharmacology principles and pharmacogenetics. So there are some people who want it. There's quite a few other people that would like to ask you questions and take the prerogative of the moderator and say that we'll please hold those questions for the discussion at the end because we want to hear our next speaker, Laurie Cavallari from University of Florida. And Laurie's going to be talking about evaluating outcomes with genotype-guided anti-platelet therapy.