 Thank you. It's great to be here and have a chance to tell you all what we're doing at Chicago. This is a screenshot of our brochure, the 1200 Patients Project, which we call the Future of Personalized Medicine. It's obviously not the future of everybody in this room, but I think for the vast majority of patients and physicians it is. And this is the flagship project for our new Center for Personalized Therapeutics. A number of people are involved in the Center, Peter O'Donnell, who's here, and as well as Nancy Cox. And this is really Peter's study. Peter should really be up here presenting. And he's the principal investigator of this protocol and has really done all of the heavy lifting. So these are barriers to implementation of pharmacogenomic diagnostics. I think we're all familiar with this, and I really want to talk about the solutions. So let's start with lack of MD knowledge. And we've created a web portal to solve this problem. Availability of tests, and we plan, and we're planning to genotype all relevant pharmacogenomic variants. What's relevant? Okay. And the question, the answer is it's relevant if an expert in pharmacogenomics would be willing to tell a consultant, a physician who consults you, that a patient has particular polymorphism and what it means. That's what the definition of relevant is. It's not something that's been demonstrated in randomized clinical trials. It's something that an expert would be willing to put in the medical records under their signature or the equivalent thereof and say, your patient has this particular genotype and this means the following for your patient. So costs and reimbursements, there will be no marginal costs for the actual genotyping. This is preemptive genotyping, so we have no delays. And the most important thing is to deal with the MD concerns, which is we're going to be providing individualized virtual pharmacogenomic consults regarding every patient and every medication. So that's the system. And this is the way the study looks. In the summer of 2010, we started working on the protocol. Peter identified what we called our early adopters. We wanted to do our, this is a phase one study. I've got a long history of doing phase one trials of new drugs. This is a phase one study of a new system. And we wanted to identify the best case scenario, a subset of our physicians in the Department of Medicine. We picked 13 across the department as I'll show you in a minute. The protocol was approved by our IRB in January of this year. And we plan to begin genotyping once we have 400 samples in hand and a contract with a CLIA lab. We have the 400 samples. We don't have a contract yet, but we hope to be genotyping over the winter. And then ideally we would like to go live in return results have a pharmacogenomic consult available beginning and about February. And then at some point in the future, we will genotype the remaining patients and we haven't decided whether we're going to do two batches of 400 or a single batch of 800. And really we're, this is a phase one study. We're trying to figure out how to do this. And we'll be examining all kinds of things, including whether the patients and their physicians even remember they're in the study when they come back to see their physicians. Here are the 13 physicians. And you can see it encompasses a large number of types of physicians. We have primary care physicians, both young and old. We have pulmonologists, hepatologists, gastroenterologists. We have people that have been using pharmacogenomic tests and are interested in people that have never ordered such a test in their life. And the eligibility are quite straightforward. It must be adults receiving ongoing outpatient care from one of our co-investigator physicians. They must be on at least one regular use medication but not more than six, unless you're over 65 years of age or expect to require a drug within five years. In other words, we sort of assume that as one ages one is at risk of taking medications. And therefore if you're on no medications and you're over 65 and you have the right physician, you can get in the study. We are excluding patients with severe life-threatening disease, i.e., a life expectancy of less than three years. We're not doing patients with metastatic cancer, for example. And we're not including patients that don't have their own organs. The study was approved by the IRB January 14th. The first patient was consented six days later, and the first patient's sample was collected five days after that. And we've now approached 592 patients. Only 16 have declined, 576 signed consent, eight ended up withdrawing consent, and of the 568 consent and patients, we have collected samples on 440 of them. And here's the drugs that they're taking. And these are all the medications that are being taken by at least 5% of patients, the most common one being hydrochlorothiazide. It's a lot of the stuff that patients receive in the primary care setting, drugs for hypertension, thyroid replacement, cholesterol, prednisone, warfarin, azathioprine is one of our primary care physicians that I prescribe all those drugs except for azathioprine. So these are common drugs. And when we went through and looked at whether there was evidence that we would consider sufficient in quality to present back to these physicians, there is, in our opinion, for all of these drugs. So they were very happy to hear that. This is what it looks like, and I'm sure this will change before we go live, but we have a physician portal. This is not linked to our EMR. This is a research portal. And the only thing in the research database, the only data are the genotype data and the medications that the patients are on. Those are the only two types of data that we're putting in this research database. And each physician will obviously have their own login credentials and their own roster. And here's a patient, a mock-up patient that is in study at Rebecca Clark. And Rebecca Clark is on two medications, amlodopine and omeprazole, both of which have actionable evidence. You see amlodopine has a green light, which means it's nothing to worry about. Omeprazole has a yellow light, means that there's something you might want to be concerned about. We indicate what the level of evidence is. We borrowed this from Russ Altman. And you see amlodopine's level two, omeprazole is also level two. And then we provide the actual PubMed IDs if physicians want to go and look at the actual data. And then one can drill down. This is a different patient, Mike Smith, a different fake patient. And Mr. Smith is on Adelimumab. And you can see the green light. And right there, and if one blows this up, here's the kinds of consults. We think of this as consults. We do not think of this as labs. We do not think of these as genotypes. So it says your patient has a genotype at the TNF gene promoter that predicts the highest likelihood of response to Adelimumab based on changes in disease activity score in 28 joints at 24 weeks. We don't tell them what the gene is. We don't tell them what the genotype is. We don't confuse them with things they don't want to know. We only tell them what they care about that their patient is likely to do well on this drug or more likely than the average patient based on the existing evidence, which is only level three evidence. And they can take that, ignore it, say I was going to do it anyway. That makes me feel good. They could say I need a real pharmacogenomic consult. That's the piece of this system we haven't quite figured out. We're really successful here. They're going to want lots of real consults. So we haven't gotten to solving that problem. And this is what we think is really cool, that they can search for things that they're considering doing. So if their patient has hypercholesterolemia, they can type that in. Or if they want to prescribe a particular drug. So in this mock-up of their concern about cholesterol and it'll pop up, don't prescribe synvastatin or be concerned about prescribing synvastatin. And synvastatin comes with a yellow light and that this hypothetical patient is homozygous for the genotype that confers the increased risk of synvastatin myopathy. So this is what they can do by searching. So we never prescribe the results. Of course an expert could reverse engineer the genotypes. But we're not burdening them with actual genes and genotypes. So here's where we are. We're getting ready to hopefully genotype in the next couple months where hopefully we'll be going live soon thereafter. And we'll hopefully have a lot of data a year from now if not sooner. And as far as opportunities for collaboration, this is a phase one trial. Phase one trials are usually followed by phase two trials if you can figure out how to get it to work. And although I have been outspoken about not wanting to do randomized clinical trials for each and every drug gene interaction, I think in this context of a system, one could think about a randomized trial. And it's really quite straightforward. One could just genotype all of the patients in the study, but then only provide the information back on half the patients. And then, of course, you can analyze the adverse events and any other outcomes that would be associated in both arms. And so I think this is, you know, this is one way one could move forward. And we're obviously going to be running this phase one trial for a while, but it's certainly not too soon to start planning the next phase of this investigation. So I look forward to your questions. Very much Mark. Let's go with, back here if we would. Okay. And then we'll get to you Mark. Do you have any mechanisms or plans for dealing with updates to the published knowledge? Yes. We've absolutely considered that. And we do plan to update, do some update genotyping on some regular basis. We haven't figured out exactly what that will be and how we'll do it because we're still working out exactly what platform we're going to use. But one would envision that one would do updates right now, probably every couple of years. That's what I mean. Yes. No, no, we don't need to do that. So first of all, we already have the DNA samples. So there's no reason to re-contact anybody. So, but we can push, if there's new information on things we've already genotyped, we just push it right out at them and make it clear that there's new information. If there's new variants that we want to genotype that we have in genotype, then we would, we're planning to do some update genotyping, as I said, every couple years at this point. One of the issues in designing this kind of study is, of course, you want enough instances where the patient actually takes something that you have data on. So I'm wondering how you picked the number 1,200 was a purely an economic issue and how did you pick the types of patients you included to try to make sure that you have enough interactions between their genotype and the medicines that they're likely to get prescribed? Well, this is just a phase one study. And we're really just testing feasibility to even see if we can get anybody to look at the information there. So step one, the physician has to remember that they're patients on the study. So we're, we've got to do that because we're not linked to the EMR. But then we just even have to get a sense as to whether they're looking at this, whether they're finding it useful, whether they're enrolling more patients on the study over time. And so the 1200 was just an arbitrary number. It was largely an economic decision. When we go to the phase two study, we're going to need to have real sample size and real statistical end points. And having this phase one study will help us to figure out how many patients, how many events we're going to need. So you mentioned that the clinical outcomes are related to genotype associated adverse events. And I think one of the concerns that I've had is that there are a number of these variants that have impacts not only on adverse events, but also on efficacy. And sometimes they're the flip side of the same coin. You know, so I mean, one way to avoid an adverse event associated with a drug is not to give the drug. In fact, that's the best way to avoid the adverse result. But if there's an efficacy thing, are you going to be able to, or do you anticipate trying to give clinicians or patients the idea that, hey, there's, you know, you need to consider both the efficacy and the adverse events to really make an informed decision about whether to use this drug at this dose or not? We're giving them, we'll give them all the information. So I mean, you know, if they, if we have CYP2D6 information and they're receiving Tamoxifen, we'll give them all kinds of information. And as far as this is just an example of one of the outcomes one might do in a phase. So I'm, we're not committed to any design. That's just, you know, a for instance, one could look at this would be a lot harder to look at efficacy outcomes across a whole bunch of diseases. But if one we're dealing with some really common diseases, one could look at cholesterol or things like that where one could measure efficacy outcomes. So I was just wondering if in this randomized design where your genotype everyone and only return the information for half whether you would anticipate getting any pushback from your IRB about having information that's medically relevant potentially and not sharing it. I suspect my IRB would not allow that. The answer is a physician can always order a test. We're not preventing them from ordering a test. So I mean, that would actually be one of the confounders in the study that if physicians were so enthusiastic about what they were doing and they say, Oh, I wish I had a CYP2D6 genotype on this patient, I'll just go and order it. We wouldn't restrict them from doing it. But so no, and as far as our IRB is pretty progressive. And I think the other point is this wouldn't have to be hidden forever, it would be hidden for whatever period of time we felt we needed to address the primary endpoint. Yeah, I am not sure I would characterize it as progressive to say if you have clinically relevant information for a patient that you wouldn't necessarily share it. I think it's a bigger problem potentially. And your IRB may allow it, but I'm not sure every IRB would allow it. And I would, you can say a physician can order it, but a physician can order a BRCA1 and 2, but they generally don't. But if I know a research subject has it, I get that information back to them. So I just think there's some issue there. Well, we are not doing any variants that predict for risk of disease. We're excluding that. So we're only doing pharmacogenomic variants. Do you consider at any point a patient portal? My understanding is there's just a physician portal. Obviously would, you know, complicate phase two concept, but particularly just remembering on both the physician side as well as the patient side that they're in the study. Yes, we are definitely considering a patient portal, but we haven't, we haven't had the time to develop that yet, but that we definitely see that as being of interest. And we're going to ask the patients if they'd like something like that. But one closing remark from Mark, Mark, did you think there was someone else? Okay, so and then just before before he goes, then lunch will be just outside and actually I'll make an announcement about that. So two things related to Gail's point. One is let's not confuse clinical plausibility with actual proven utility. And I think that one of the problems is we have these really elegant mechanisms by which we think this is going to work, but we have innumerable instances in medicine where things don't work the way we predict they're going to work. So if we really don't know the answer, we have to be able to do studies to get at the answer. And so I would support that. I think there are different ways that could potentially be done. One one way is to put the option to order the test within the clinician workflow and then look at the outcomes and patients in which clinicians order the test versus not order the test and, you know, do randomization at that type of a level. Now again, these are messy. The the results are going to be more difficult to interpret. But you know, we can't just blithely assume that we're going to have the impact that we think we're going to have because that's just not we're just not that simple as biological organisms. Right. I mean, I think some of us in this room have spent a lot of time doing randomized clinical trials of drugs and Dan Rodin probably has the most high profile experience of doing a trial with a drug that everybody was sure was going to work and didn't, which was fleckonite. So so I think that, huh? So I think we have to be very careful of not believing our our own stuff too far. And yeah, then then it should be required for any test that's actually required that it be ordered before drug is prescribed. I'm all for having it be done. We don't all define know the same way. It's getting good getting rid of patients too. So we do. It's just as an incidental luck. We do we do need to finish up so