 So thank you for inviting me here to speak today. Can you hear me okay? Okay. So I'll be presenting on behalf of UF and also on behalf of the Ignite Pharmacogenetics Working Group. So we heard earlier from Laura on the Emerge Working Group, and so I'll be presenting work from the Ignite Group. So this slide shows our timeline for pharmacogenetic implementation at UF Health. And as you can see, our initial implementation was with SIP-2C-19 testing to predict response to clopidogrel and guide antiplatelet therapy after PCI. And we chose to start with this particular implementation for several reasons. One is that there is very consistent evidence that there is decreased effectiveness of clopidogrel in individuals with a loss of function SIP-2C-19 variant. There is now a box warning on the labeling for clopidogrel, warning about reduced effectiveness of poor metabolizers. And then of course we have CPIC guidelines available to assist us with translating the genotype results into prescribing decisions, and they specifically recommend the use of alternative antiplatelet therapy, usually press and grill or ticagrel or in the absence of any contraindications to individuals with a loss of function variant. About a year after we started our implementation, we received funding from Ignite as part of that network. And then since then we've really expanded what we've offered. We started implementing SIP-2C-19 testing for clopidogrel in our Jacksonville campus about a year ago. Here we're using a rapid genotyping platform, so it's a little bit different from what we're doing in Gainesville. And then we've also implemented a number of other gene drug pairs into practice. We recognize that it's important with many of these that we contribute to the evidence based on outcomes, as we talked about yesterday. And so with many of our implementations we have designed for pragmatic studies around these. And we also talked about yesterday how it's hard to do these at a single institution. So we have collaborated with other institutions to look at outcomes. For example, we're working with Moffitt and Howard just down the street from us to look at outcomes with SIP-2D6 testing to guide opioid prescribing for pain. And then we're working with DeMores Children's Hospital in Orlando to look at the effect of SIP-2C-19 guided PPI dosing on GERD and dyspepsia symptoms. And we have a simultaneous study looking at computable phenotypes related to adverse effects with PPI's with Vanderbilt and SIP-2C-19 genotype. But I'm going to focus from here on out on our work with SIP-2C-19 and Clopidogrel. So just very briefly, Clopidogrel is a pro-drug. Once it enters the body, reaches the liver and undergoes this two-step process to bioactivation. And SIP-2C-19 is involved in both steps of this bioactivation process, leading to this active metabolite that then binds irreversibly to the P2Y12 receptor to inhibit platelet aggregation. We know that about 25 to 35 percent of whites and blacks and up to 65 percent of Asians have a non-functional or loss-of-function variant. I'll be calling it LOF or loss-of-function variant for the rest of the talk. Sorry, Mary. I know that's not exactly per the standardization, but that's how we refer to it when I presented these data at AHA, and I'll be showing those slides here. So in these individuals, they may have reduced production of active metabolite and decreased inhibition of platelet aggregation with Clopidogrel. There are, as I mentioned earlier, consistent data mostly from retrospective or post-talk analysis of randomized controlled trials and patient registries that have shown decreased effectiveness of Clopidogrel in individuals with a loss-of-function allele. In one of the more highly cited studies, it's a meta-analysis by Jessica Meaghan or colleagues where they looked at nine different trials, 9,700 Clopidogrel-treated patients. They were very high-risk patients, most of whom had an acute-coordinary syndrome and underwent PCI, and what they found is that the risk for a major adverse cardiovascular events, which they defined as CB death and Mayan stroke, in addition to stent thrombosis, was significantly greater in Clopidogrel-treated patients with versus without a loss-of-function variant. There is an ongoing randomized controlled trial that's looking at clinical utility with genotype-guided antiplatelet therapy after PCI, and it's called Taylor PCI. It started in 2013, and they're targeting about 5,300 patients who undergo PCI, and they're being randomized to a genotype-guided strategy where ticagrelor is used for loss-of-function allele carriers and Clopidogrel is used for others versus Clopidogrel without genotyping. And they're looking at MACE at one year, but unfortunately we won't have the results from this for another three years. So when we started our currently, with our implementation at UF Health, what we found is the most feasible approach for providing genotype-guided therapy is to place the genotype order on the post-PCI order set, so that way, and it's actually defaulted, so that every patient who undergoes PCI will get genotyped, unless the physician chooses to deselect that test. We run the test in our pathology laboratory. It's actually located a bit off campus, and there's about a two to three day turnaround time, and then the result, once it's done, it's placed in the electronic medical record. We have a clinical pharmacist that then follows up on all of the genotyping results, and for patients who have a loss-of-function allele will notify the treating physician about that and recommend alternative therapy as long as there are no contraindications. And then we built clinical decision support, which basically, if clopidogrel is prescribed for a patient with a loss-of-function allele, then an alert will appear stating that there is a risk for reduced effectiveness in recommending alternative therapy, and similar to what Josh Denney showed yesterday, it has the contraindications to these drugs and other information that maybe needed to really smooth this prescribing process and make it very simple for the prescriber. So we implemented this as part of clinical care. This was not an a priori design clinical trial, but after the first two years of the program, we decided to go back and see what happened with our patients who received genotyping after PCI. So we reviewed the medical records. We had 408 patients at this time point, and we collected data on re-hospitalizations for adverse cardiovascular events, including myocardial infarction, stroke or synthrombosis, as well as cardiovascular death that occurred within six months of them getting a PCI in genotyping. We then compared this composite MACE endpoint between patients who had a loss-of-function allele and were treated with alternative therapy, this predominantly consisted of ticagrel or oproxagrel, versus those with a loss-of-function allele treated with clopidogrel. So these are initial results from UF Health. This is the cumulative MACE rate over six months after PCI. And what you can see, that top line here where it's red, those are patients with a loss-of-function allele treated with clopidogrel, and you can see they have a much higher risk for MACE compared to those with a loss-of-function allele here at the bottom on alternative therapy with the blue line. And there's no difference between those who had a loss-of-function allele treated with alternative therapy versus those without a loss-of-function allele. And again, this was in about 408 patients. About 30, I think it was 31% had a loss-of-function variant. So then we decided we wanted to look at this in a much larger group, and we did this within the Ignite Pharmacogenetics Working Group. So the working group consists of sites that are funded within the Ignite Network, in addition to a number of affiliate members. We have, I think, at least 10 affiliate members in the Pharmacogenetics Working Group. Most affiliates are in the working group. And we do have a little paper. I don't have the reference here, but we have some information and overview sort of of the working group and what we're doing that was published in Clinical Translational Science. I think it's either in press or online ahead of print or was recently published. So of all of the groups in the working group, the primary implementation was CYP-2C19 testing for clopetagrel. And so we decided that we would look at outcomes with this approach to anti-platelet therapy. So the seven groups shown here in blue contributed data to this analysis, and then the two other groups contributed data on contemporary use patterns with anti-platelet therapy. I won't show that here, but basically they showed that clopetagrel is still the primary anti-platelet drug that is used after PCI, even in patients with an acute coronary syndrome. So this is a multi-center pragmatic investigation of outcomes. All seven sites involved, as I mentioned, had clinically implemented testing into practice. So this was not done as any part of a clinical trial to guide anti-platelet therapy after PCI. At each site, alternative therapy was recommended in those who had a loss of functional allele, but the ultimate prescribing decision was left up to the physician. And there was no genotype-guided recommendation made for those without a loss of functional allele. So our aim was that among these patients who got tested at the time of PCI to compare the risk for MACE, or major adverse cardiovascular events, between those with a loss of functional allele on alternative therapy, which is predominantly Prasigrel or Tychagrelor. And we'll call this the LOF Alternative Group, versus those with a loss of functional allele on standard dose clopetagrel. And I'll call this the LOF Clopetagrel Group. And then we also compared MACE between the LOF Alternative Group and patients without a loss of functional allele, which I'll call the non-LOF group. So also in sites manually abstracted data from the electronic medical record, we talked yesterday about coming up with a common, you know, defining common outcomes to look at. And so we actually spent a good amount of time with everyone who participated in defining the data elements that we wanted to collect, and specifically defining how or determining how we wanted to define MACE. We came up with this common data collection sheet. We used the data collection sheet from the International Clopetagrel Pharmacogenetics Consortium as a baseline to start with. And then we went from there. I think we ended up with about 60 variables that we collected, clinical variables, variables related to the outcomes. And this data collection tool is actually posted on the Ignite website in the toolbox for others to use to examine outcomes. So each site reviewed the medical records for their patients. They collected data on medication use, specifically on antiplatelet drug therapy, clinical variables, hospitalizations for cardiovascular events, and deaths that occurred over a 12-month period after the PCI when they got the genotyping done. And then all the data was sent to UF where it was curated, and then it was aggregated together for analysis. So our primary outcome was MACE, and we defined this as a composite of death MI or stroke within 12 months after the indexed PCI. That was the PCI associated with genetic testing. We assessed antiplatelet therapy at the time of the event or the last encounter for those who did not have an event, and for individuals. And this was the time of last encounter when the antiplatelet therapy was documented. And we also looked at time to prescribing antiplatelet therapy in individuals with a loss of functional yield. We used Kaplan-Meier analysis to compare time to MACE between the groups, the LOF alternative versus LOF clopetagro, and LOF alternative versus non-LOF group. We used logistic regression to summarize the clinical differences between groups into a propensity score, and then the hazard rates were compared with Cox regression analysis with propensity score adjustment. So these are the results, and we presented these at the American Heart Association this past November, and the manuscript is currently under review. And I forgot to mention that the UF health data we had actually presented the previous year at American Heart. So we had 1,815 patients in the total cohort. 31.5 had a loss of functional yield, which is consistent with what's reported. Of these, 60.5% were treated with alternative therapy. And then if you look in comparison among patients without an LOF allele, only 15.5 were treated with alternative therapy. So this was a significant difference in the use of alternative antiplatelet therapy between those with and without a loss of function allele. And the predominant antiplatelet therapy that was used was Prasegral. It comprised over 60% of the alternative antiplatelet therapy used. These are the patient characteristics shown by the LOF alternative, LOF clopidogrel, and non-LOF groups. So overall, the mean age was about 63 years. We had mostly men and mostly people of European ancestry. This was a very high-risk population. We had about a 38 overall prevalence of diabetes. And about 67% of individuals had an acute coronary syndrome at the time of their PCIs, a very high-risk population. There were differences between groups, which is expected. This is clinical care. This was not done in any of the sites as a clinical trial. You can see that compared to the LOF alternative group, those in the other two groups were slightly older, had a higher prevalence of diabetes. There was a higher greater history of struggle TIA in the LOF clopidogrel group, which likely led to them using clopidogrel versus alternative therapy. And then I think there were more whites in the non-LOF group. Other differences were similar between groups. In terms of implementation metrics, so again, this was done as clinical care, but very nicely surprisingly, I guess, there was very efficient implementation of genotype-guided therapy. So the median time to genotype among these groups was one day, and the median time to alternative anti-cladelet therapy in those with a loss of functional yield was also one day. This is the results. So we saw a significantly higher rate of maze in this LOF clopidogrel group at the top compared to the LOF alternative group. There's the log-ranked p-value of .016. There was no difference between the LOF alternative group in maze versus those without an LOF allele. Again, most of these individuals were treated with clopidogrel in the non-LOF group. And you see the log-ranked p-value there. And then when we did Cox regression analysis with propensity score adjustment, there remained a significantly higher risk for maze in the LOF clopidogrel group versus the LOF alternative group with a hazard ratio of 2.21, and then no difference between the LOF alternative and the non-LOF groups. So our conclusion based on these data, one, we showed that this genotype-guided approach to therapy is feasible across multiple institutions in the U.S., and that we had very efficient return of genotype results overall and institution of alternative anti-cladelet therapy. And then our data suggests that in patients who have an LOF allele, we can improve cardiovascular events if the genotype is made available early after PCI and we implement alternative therapy in these patients. So in terms of next steps, one of the things we talked about yesterday was the importance of economic data. So we are working on an economic analysis. This is being led by Josh Peterson from Vanderbilt and Needle-Empty from UAB. Josh has a simulation model, which Josh Denny presented yesterday. And the preliminary work with that model has been completed. We're now refining it a little bit and doing some sensitivity analysis. But basically, we have a base case of clopidogrel for everyone, and that's compared with a approach of alternative therapy for everyone without genotyping. Genotyping at the time of PCI with alternative therapy used in those with a loss of function allele, others getting clopidogrel, or use of alternative therapy for the initial month in all patients. The genotype determined in having that available at one month and to guide antiplatelet therapy from that point on, so that those without a loss of function allele could be switched to clopidogrel. So be looking for those data soon. And I also want to mention that final approach with using alternative therapy first is what our UF Health Jacksonville site is really using for most of their patients. So we see a little bit of practice pattern change even though clopidogrel remains the primary drug. We are seeing a little bit of uptake with alternative antiplatelet therapy, and so this cost-effectiveness analysis will hopefully be very contemporary and help address that or take into account that. And then we're also working on a manuscript. This is being led by Phil Impey from Pittsburgh describing the actual strategies for implementation among the sites that were involved in addition to other sites to really showing there are more than one approach to implementing this into practice, but it still leads to improved outcomes. And I just wanted to conclude by acknowledging all the investigators that contributed to this, including Julie, our PI, and Kristin from UF, Todd Skar, who's here, Dan Rodin, Josh Denney, and then the rest of the investigators here, and then our significant funding from NHGRI and other institutes for this work and other sources as well, which are listed here at the bottom. Thank you. Thank you. I think we have time for one or two clarifying questions. In addition to PCI patients, clopidogoris, you see a number of other clinical categories. Would you like to comment what did you use PCI patients and would it genomic guided therapy work in other patients' categories? So we specifically focused on individuals undergoing PCI because the data that are available suggests that the risk for adverse outcomes in those with a loss of function allele is really those very high-risk patients. Such as those undergoing PCI. The data in other populations, such as those with stable coronary disease or individuals with ACS who are managed medically without PCI are more conflicting or inconsistent, more inconsistent than conflicting, I guess. But overall, the data consistently support an association in PCI patients, but not in other groups. What is, would be the proportion of PCI categories subpopulation to other subpopulation which also are given clopidogoris? So all of the patients here were PCI patients in our analysis. But clopidogoril is widely prescribed beyond just PCI, you're right. But the data supporting genotype guided therapy is really that PCI population. We are seeing at UF Health that they're using it or that post-stroke. They're actually starting to do some genotyping for CYP2C19 to help guide antiplatelet therapy and that area is a little bit more controversial because Prasagril, for example, is contraindicated in patients who have a history of stroke, but they are using it. We actually had one patient, so Josh Denny gave this example of a patient who had five stents. We've had a patient who I think had three strokes. Each time on clopidogoril, they finally genotyped that patient in the patient support metabolizer. So there could be some implications in other areas. We just don't have the accumulation of data. Yes, Lynn? Yeah, Laurie, a beautiful presentation. Just to clarify, in terms of your MACE events according to your metabolizer status, so did you find that in your poor metabolizers they were more likely to have a stent thrombosis whereas an MI or are there correlations with the metabolizer status? That was a great question. So overall, I think about 50, I looked this up this morning so I can remember the numbers, little over 50 were poor metabolizers in the multi-center study. The remainder of LOF allele carriers were intermediate metabolizers. Almost all of our poor metabolizers were treated with alternative therapy, so the clinicians are paying attention there and maybe the box morning helps. However, among the intermediate metabolizers, that's where we saw the use of clopidogoril on about 40% of those who are intermediate metabolizers versus alternative therapy. So we actually did do a sort of subanalysis in that group looking at the composite outcome. I don't recall with the individual events, but we still saw significantly fewer events and intermediate metabolizers treated with alternative therapy versus intermediate metabolizers treated with clopidogoril and actually someone from the FDA has already contacted us to say what does it look like in your intermediate metabolizer? So perhaps this can help to revise the labeling for clopidogoril. Because I think that is really critical that the intermediate metabolizers are also experiencing MACE at a high rate. Yeah, I agree. So it's a large portion of the population. It's 30% who have a loss of function variant so it has large implications. Okay, Bob, I'm gonna ask that you hold your question for the discussion. Laurie, thank you for a great presentation and now we have Mark Williams who's gonna be talking to us about pragmatic applications of economic and cost-effectiveness analysis.