 All right, well thank you so much for that introduction, Joji, and I guess now it's officially good afternoon everyone. Presenting at THI Grand Rounds is really an honor for anyone, but it's particularly special for me because just six years ago I was a fellow in the audience watching folks lecture up here. So this is really an honor for me. I wanted to just up front thank Doctors Willerson and Colter for not only inviting me to speak today, but for also what I consider a joyous family reunion that happens to be perfectly timed around Thanksgiving. So many thanks to them. Like any good family reunion, I think it's pretty important to reminisce about the past before we get to present business. So first and foremost I'd like to acknowledge the visionary leaders that made this place and my training here specifically possible, the late Dr. Cooley and Dr. Willerson. During my three years here, Doctors Colter and Stainback shared their incredible knowledge of echo and clinical cardiology with me, and we published some really great papers together. So I just wanted to say thank you to both of them for their mentorship. One last thing, there are the boots. So these just aren't any boots. They're the only ones in the world that are laser engraved with the THI logo thanks to Joji and Steve. And I'm proud to say that I wore those boots traveling to Houston yesterday to dinner with Joji last night, and I'm absolutely wearing them right now. So thanks Joji and thanks Steve. All right, so I think that's enough reminiscing, though I do promise a few more good pictures during the talk. As great as it was being here at THI, as Joji mentioned after fellowship in 2013, I went to the Brigham and Women's Hospital in Boston to learn from this man, Marcelo de Carly. Today, I want to share with you what I learned from him, the research that we conducted together and how we think our findings will help drive new paradigms to treat coronary artery disease. So toward that end, here are the topics that I'd like to cover in my talk today. First, talking a little bit about the quantitation of absolute myocardial blood flow and coronary flow reserve or CFR, as I'll mention to it during the talk, from dynamic myocardial perfusion pet images. Second, we'll talk a little bit about how CFR compares to invasive fractional flow reserve measurements and why there's an expected imperfect correlation between these measures. Next, I want to talk about CFR as a powerful independent predictor of adverse cardiovascular outcomes and as a potentially useful tool to guide clinical management. And then I'll just give you a sense at least at the end about where we're going to go next with this information from transitioning from what we're calling coronary flow reserve to coronary flow capacity and how that might even improve what we've done so far. So, you know, when I arrived in Boston, Dr. DeCarly had already established a robust pet myocardial perfusion imaging program, and I'm glad to see actually coming here today that a similar program was started in 2013 by doctors Moran Ford here. One of the unique advantages of pet MPI over traditional SPECT MPI is that you can acquire these dynamic images that you see here with relatively high spatial resolution. And just to kind of orient you to what you're seeing there at the top of the screen. So you've got either rubidium 82 or N13 ammonia as your radio tracer, and you're injecting it into the blood pool, and that's what you see at the beginning of the loop. As time goes on, that tracer goes from the blood pool into the actual myocardium. And it turns out that what you're seeing qualitatively here can be translated into quantitative data that you see on the right that I'm going to highlight on the next slide. So what you're seeing on the right hand side of the slide with those graphs in green is the tracer activity in the blood pool, which as you can expect has a spike very early on at both rest and stress. But as time goes on, that tracer then falls out of the blood pool and gets into the myocardium as we saw on those qualitative images before. Well, these time activity curves can actually be made for each segment of the LV myocardium. And as you see on the left here, that can actually be turned into absolute myocardial blood flow in terms of milliliters per minute per gram of myocardial tissue. And you can do that not only over the entire LV, which is designated by the TOT line at the bottom, but you can actually do that for each specific coronary artery territory, the LAD, the left circumflex, and the RCA. So coronary flow reserve, which is the last column that you see on the right, is really your absolute stress flow divided by your absolute rest flow or stress over rest. So Dr. Moore is in the audience, and he and I and Dr. Ford probably think that's pretty awesome on its own, and thank you for coming, Dr. Moore, to this talk. But Steve, the Scotchmaster jogger, isn't as easily impressed. As a true interventionist and as an acolyte of Dr. Fish, he doesn't really believe anything unless it's been vetted with a catheter. So let's see if we can convince Steve and Dr. Fish. So these two studies, these two different graphs are from two different studies, but they show essentially the same thing. The graph on the left is really sort of, and on the right, are showing mycardial fractional flow reserve or CFR versus, sorry, mycardial fractional flow reserve as measured by a catheter versus relative flow reserve by PET. And basically, what they're showing you is that there is a good correlation between the invasive measures and the noninvasive measures by PET, but as you can tell from these graphs, the R values really aren't fantastic. It's 0.1 to 0.87 in these two separate studies. So there's excellent but not perfect correlation, and I want to sort of explain to you why that's not unexpected based in part on work that was done by land school just down the street. So as shown here by Dr. Gould, a discrete epicardial coronary artery stenosis of greater than 70% as you might expect typically causes both an abnormal invasive FFR and an abnormal noninvasive CFR. But if you don't have a discrete stenosis and instead if you have diffuse narrowing, then what you're going to see is that the CFR by PET can be reduced, but there's only a minimal fall in segmental pressure gradient or FFR for any segment of the artery because there really is no discrete stenosis that's causing a specific discrete pressure drop. And so this slide also by Dr. Gould goes a little bit further and I think is really sort of explanatory about why CFR by PET is truly powerful. So it not only integrates, so on the top you've got blood flow in a normal artery and a CFR of four. Now in the second diagram, you've got that scenario where you've got a discrete epicardial stenosis and both of them, the CFR and the FFR would be abnormal. In the third depiction, you've got the CFR showing that it's abnormal, but the stenosis may not actually be hemodynamically significant and may not cause a drop when invasively measuring by FFR. So that's all what I've told you before as well and this is just a different way of depicting that. As far as diffuse disease like in number four, CFR will pick that up, but FFR won't because again there's no pressure drop. Now what's really interesting is that the difference between five and six is arterial remodeling. So CFR can pick up the difference between number five and number six where six is the artery remodeling and the CFR is actually relatively preserved whereas without the arterial remodeling in number five, it's still low. And so that's the real advantage of CFR as far as the epicardial coronary arteries are concerned. But it integrates one more thing and that's really seen in this remarkable pathology slide which is that the microvascular actually comprises the bulk of the vascular volume and the resistance in the myocardium, not the epicardial vessels. For the past 50 plus years, we've focused the vast majority of our research in terms of how to diagnose and treat epicardial coronary disease. Petriive CFR on the other hand incorporates the hemodynamic effect of all of this microvascular disease that might be occurring as well across the entire LV and therefore has the potential to radically shift our thinking about the prognosis and treatment of coronary disease and I want to show you now some data to back up that claim. So these are some seminal data by my colleague, Venk Murthy and Dr. DeCarly that were published before I arrived in Boston. And these data are really the first to show that diffuse atherosclerosis and microvascular disease have an independent and incremental effect on outcomes and patients with CAD. So just to walk you through the slide, on the x-axis you've got the degree of ischemia, on the y-axis you've got annualized mortality and then turtiles of CFR from the upper to the lower turtiles. So as we kind of all know from just regular pet perfusion or spec perfusion, the degree of ischemia that you have on relative perfusion images determines whether or not you have higher risk or lower risk. But what's really interesting about this slide is the people that have zero ischemia, zero percent ischemia, so just look at that column. If you're in the lower turtile of CFR versus the upper turtile, the difference in risk is absolutely impressive. And so people that don't have ischemia by traditional perfusion measurements might actually still have risk. And we think that that may be because they either have diffuse atherosclerosis or they have microvascular disease that's not really as prominent in their epicardial vessels. So in extended follow-up over three years, there's a clear difference in cardiac mortality between all three turtiles of CFR. And this is even after adjustment for age, sex, diabetes, LV ejection fraction, some stress score, which is a summary measure of ischemia and scar on myocardial perfusion imaging and other covariates. So you can see from here that there's actually a five-fold difference in mortality between the lower and upper turtiles of CFR, even as you follow these patients out for three years. So finally, and most importantly, so that's nice from a prognostic perspective, but from a clinical perspective, why does CFR matter? Well, these are net reclassification data that really show that CFR adds incremental value to traditional risk stratification, including clinical risk factors, LV ejection fraction and some stress score. So on this slide, I want to point you most importantly to that intermediate risk group, which is the group that sort of is the hardest to understand and test and treat. So when you look at those people who are intermediate risk by traditional risk scoring, if you incorporate CFR, about 50% of those people can be net reclassified into low risk or high risk, and only 50% of them stay in intermediate risk. And most importantly, if that group that's in red is down classified, then maybe we don't necessarily have to treat them the exact same way as we do the other folks who are truly intermediate or high risk. And that's on the order of 17, 18% of folks. So, bank and doctor to Carly next focused on the prognostic value of CFR in what we traditionally call a high risk population, patients with diabetes. So this slide demonstrates their findings. So what I want to sort of do is move you from left to right. So folks that have diabetes and have established coronary disease obviously have the worst outcomes in terms of cardiac mortality. And they look very similar to people that just have coronary disease without diabetes. Now, folks that have coronary, do not have coronary disease, but have diabetes and have a low CFR, those folks have a pretty high cardiac mortality. That's the blue bar in the middle. But interestingly, people without coronary disease with diabetes that have a preserved CFR, meaning above 1.6 in this particular data set, their risk looks exactly like people who don't have diabetes and don't have coronary disease. And so this group that we kind of traditionally assume is universally high risk is actually not so. And that's an important piece of information. And that may be at least prognostically where CFR might be able to help. So based on Vank and Dr. DeCarly's work, I thought that it was at least worth testing the same hypothesis in another so-called high risk group. And the group that I picked was dialysis dependent end stage renal disease patients. So as you can see in our cohort here, we included patients with and without overt cardiovascular disease, meaning people that had either MI or coronary revascularization with cabbage or PCI in the overt CVD group and those without any of those things in the no overt CVD group. And as you can tell, if you take sort of the median global CFR, which is considered sort of the threshold for normal, there is a huge percentage of this population that has a global CFR less than 2. But there's actually a fair amount of this population that has a global CFR greater than 2. And so if you split the cohort by the median CFR and here the cutoff was 1.4, and you'll notice that the cutoffs here for CFR are different in many different studies because we still haven't established an actual threshold that we say that we can use clinically. And that's going to take a little bit more time. But in this case, it was a CFR less than or greater than 1.4. So just like Vank found in the diabetes population, in the ESRD population, we found the same display and risk that those that had preserved CFR ended up doing better than those that had reduced CFR. And we found that more cardiovascular mortality just like Vank did. Now what's really interesting is that we found the same display and risk for all cause mortality. So obviously folks with end stage renal disease often die of cardiovascular disease. We didn't really expect this to be true for all cause mortality and it ended up being that way. So we think at least that this is pretty powerful stuff. And in addition to those sort of unadjusted curves, this is now full adjustment for age prior coronary disease, including revascularization, total ischemia and scar burden as some stress core. Even after accounting for all of that, as you can see highlighted in the red circle, CFR remained an independent predictor of all cause mortality. And again, pretty remarkable sort of that even after you account for all of those other things that this still comes out as independent. So my colleague David Cheritan and I then asked if our findings would hold up across the spectrum of chronic kidney disease. And it turns out, as you can see in this slide that they do. So the dialysis cohort on the far right is similar to the one that I had. But looking at stage 0 through stage 5 CKD, you can see that based on CFR turtile, for the most part there's a gradient in risk associated with the CFR. And so we really think that this is a robust measurement in several different patient populations that can help us prognosticate risk better than we currently do. And so in the years since I completed my advanced imaging fellowship and left for Brown, Dr. DeCarly's fellows have continued to put out similarly powerful data in numerous other patient populations. They've done it in folks with obesity, they've done it in patients with heart failure with preserved ejection fraction, and they've done it with patients even with aortic stenosis. Hopefully at this point I've got at least a few of you thinking that CFR is a precise, reproducible, quantitative imaging biomarker of cardiovascular risk. So let's see if we can get Joji on board with some even more intriguing data that suggests that CFR might actually have far more clinical utility than simply improving prognostication. I don't know if Dr. Coulter's here, but she's done some fantastic work to promote women's heart health. And I think she'll appreciate these next couple of slides. So one of my co-fellows at the Brigham Viviani Takedi has done outstanding work to show that CFR can actually be leveraged to better understand the pathophysiology of CAD. In this study, which I was fortunate to be a part of, we showed that men with low CFR tend to have, and so just focus on the far right there. So this is a whole group of folks with low CFR, meaning less than 1.6 in this cohort. On the far right, you see that men who have low CFR predominantly have an angiographic CAD prognostic index somewhere between 37 and 100, meaning that they have significant epicardial coronary stenoses. Women on the other hand have the exact opposite profile. So now on the far left, you see that the vast majority of women with low CFR actually have non-obstructive CAD or a CADPI of 0 to 19. And so in other words, CAD in women may actually be more diffuse and may more predominantly involve the microvascular as the mechanism for lowering CFR and increasing risk. And that's not really just academic. It turns out to have an incredibly powerful implication for outcomes. So as you see here, the presence or absence of obstructive epicardial disease differentiates risk in men significantly. So men are in blue, women are in red. The difference in survival curves, when you look at just the angiographic score or the CADPI, a low score in men means that you don't have as much risk. In women, it makes absolutely no difference what your score is, whether it's high or low, you have the same risk. What's true in women though is that if your CFR is low, that differentiates risk as opposed to men. So on the right, men, when you look at their CFR, it doesn't seem to matter as much on top of the angiographic epicardial disease. But in women, if you have low CFR, it's now the risk starts to play out. So that's really important data. And it suggests that maybe our treatment in men and women for CAD might need to be different in the future. So getting to the potential impact of CFR on therapy in both men and women, doctors Christiana Patel and Tim Bateman in Kansas City showed earlier this year that the effectiveness of medical therapy and early revascularization in patients with CAD may be modulated by CFR. So here you've got the spectrum of global myocardial blood flow reserve across the LV. And you can see that at around 1.7, 1.8 or so of myocardial blood flow reserve, there's a cross in these curves. What that means is is that in folks that have a CFR myocardial blood flow reserve less than 1.7 or 1.8, medical therapy is actually not, is actually inferior to early revascularization. But in folks with a myocardial blood flow reserve greater than that, the curve switch and now medical therapy is actually associated with better outcomes than early revascularization. Now we're going to, you know, today and tomorrow see some things at AHA from the ischemia trial that will help add to these data. But my point is that this is pretty provocative and it again means that maybe treatment for CAD shouldn't be uniform for everybody. And maybe CFR is one of those ways that we might be able to differentiate folks. So this is other important work by Viviani and Dr. DeCarly again. And it shows that it actually how we revascularize patients, not just whether we should, but how we do it with either PCI or cabbage might actually one day be informed by CFR as well. So the left half of this graph is really what you've seen before, right? So that when you have high CFR or preserved CFR, your risk is not particularly high. But when your CFR is low, then things start to matter. So if your CFR is low and you end up getting PCI, yes, your risk is reduced. But if you end up getting, quote, total revascularization with cabbage when your CFR is low, now your risk is significantly reduced. And so maybe in patients with low CFR, PCI may not be as effective as cabbage. And again, that may mean that the way that we offer things to folks that have low CFR and CAD may be influenced by some of these data. I'm not saying that definitively. I'm saying that at least these data are suggestive that we might need to change some of our treatment paradigms. So I've shown you some data that CFR can augment our understanding of CAD pathophysiology and may soon be incorporated into CAD treatment algorithms. Now I want to show you some data that peak myocardial blood flow. So again, not CFR, but peak myocardial blood flow from PET can be leveraged to completely shift an entire classification system. So this is the 2010 ISHLT nomenclature for cardiac allograft vasculopathy. As you can see here, between CAV 0, 1, 2, and 3, essentially all of these categories are dependent on invasive measurements from catheterization. And folks like Joji know this very well that you end up having to cath these folks over and over again if you suspect that they've got allograft vasculopathy. Only recently has this classification now added this bottom line about using some echo measurements like EF and other measurements from right heart cath to sort of add to those invasive angiography measurements. So one question that we had was can we improve on that? So Paco Bravo, another imaging cofellow of mine at the Brigham, I and others in Dr. Carly's lab sought to create a completely new noninvasive classification for cardiac allograft vasculopathy. So I'll walk you through the algorithm that we came up with. So we said, okay, you start with your myocardial perfusion PET and you actually pay attention in this case to the perfusion imaging. Does the perfusion, is it completely normal or is it abnormal? If it's normal, then the next thing that you look at is peak myocardial blood flow. Is it preserved? In which case, that's the equivalent of what we call the CAV zero, not significant. Or is it reduced? And if it's reduced, then you look at the ejection fraction, whether it's above or below 45%. If it's above 45, then you call that mild CAV or CAV one. And if it's less than 45%, we actually classified that as CAV grade two or three or moderate to severe CAV. That's the left side. The right side, in the abnormal group, if you've got single vessel disease in terms of where the perfusion abnormalities are, where multivessel disease matters. So if it's multivessel and you've got abnormal perfusion, then you just say that they've got moderate to severe allograft vasculopathy. If it's only single vessel disease, you again go back to the peak myocardial blood flow, less than 1.7 or greater than 1.7. Again, preserved, probably mild. If it's not preserved, then you again end up in that moderate to severe category. So it's nice to have an algorithm, but now you've got to prove that it actually works. So here's what our survival curves looked like. So pet CAV, either not significant or mild, pretty good prognosis. If by our algorithm you ended up in the CAV two or three category by pet, your prognosis was significantly worse. So that's nice, but it doesn't really speak to whether or not this is any better than our current algorithms or even comparable. So just for reference, this is what the curves look like when you use the invasive current standard. So again, CAV zero and one are very similar to each other and have relatively good prognosis. It's really CAV two and three by the invasive measurements that really have the worst prognosis. Now I want you to focus really on that first zero to three year follow up on the right versus what we showed. And you'll see that the numbers are almost exactly identical. So in the CAV zero and one group at three years, you might have somewhere around a freedom from mace of around 85, 90%. Whereas those in the CAV two, CAV three category are now sort of around the 40% of them have had some sort of mace event. So if you take those numbers and then transpose them, they look almost exactly like the data that we're showing for pet alone. So why does that matter? Well, if we can get away with not invasively cath all of these patients and we can get away with a noninvasive test that gives us the exact same prognostic information and might be able to drive our treatments the same way, then why wouldn't we do the noninvasive test? The only real sort of advantage of the invasive procedure at that point is that if you wanted to do a biopsy, you could do it at that time. But again, maybe biopsies should be driven by noninvasive measurement first and whether if you have a normal CFR and you have relatively little profusion defects, then maybe you don't need to do the biopsy. So we think that's pretty exciting. So what's the next evolution? And this is the last thing that I'll talk about and then I'd be happy to take questions. So although what I've shown you is certainly exciting and important, I think and Dr. De Carly thinks and actually Dr. Gould thinks that we might actually still just be at the very beginning. These data are from Angkor Gupta, yet another one of Dr. De Carly's fellows. And this is really important data. It shows that combining CFR with stress myocardial blood flow actually differentiates risk better than either one alone. And so just looking at the graph on the left, the four quadrants. So in red, you've got low stress myocardial blood flow and low coronary flow reserve. It turns out, as you see on the right, that that red group is the group that actually has the highest risk. And if you move over to the yellow group, which based on all the other data that I was telling you before, they have low CFR. So we would have said before that they have universally the same risk as the people in red, but it turns out that they don't. And the reason that they don't is because they have preservation of the ability to augment their flow at stress. And so their coronary health in those folks is probably better than the folks in red. And it turns out that they actually have significantly less risk than the people in red. So even amongst what I've shown you, there is a differential in risk. What's also really interesting to me is that amongst people who have, quote, normal coronary flow reserve, the same thing is actually true. The preservation of your ability to maximize your blood flow at stress, meaning now the green versus the red boxes and the green versus the blue, and the green versus the blue on the right, there's still actually a significant difference in risk. So we think at least that all of the data that I've shown you is important, but we may still have quite a ways to go in terms of truly understanding pet CFR, myocardial blood flow in absolute terms, and its importance on risk. So this is a last slide again from Dr. Gould down the street, really showing that coronary flow capacity, again that's now integrating coronary flow reserve and myocardial blood flow at stress. Actually that's not theoretical either from a prognostic standpoint. We're starting to now see the same data that I showed you before in terms of treatment, now seeing it again with coronary flow capacity. So on the top, basically these are people with severely reduced coronary flow capacity. So and that's defined by CFR less than 1.27 and stress perfusion less than 0.83 cc per minute per gram of tissue. So you can see that in those people with severely reduced coronary flow capacity, PCI or cabbage within 90 days of the pet seems to be better than not doing a revascularization. But the exact opposite is essentially true in patients that have mild to moderately reduced coronary flow capacity. In fact there's a non-statistically significant trend for their outcomes to be worse if they do get revascularization. And so again we're starting to get a little bit better in terms of our risk differentiation and it may have huge impact on terms of our therapies. So in summary, dynamic images from myocardial perfusion pet offer the opportunity to quantify absolute blood flows across the entire LV. Pet derived CFR is a non-invasive precise reproducible well validated measure of total coronary health. In the absence of obstructive epicardial CAD it uniquely allows us to quantitate diffuse atherosclerosis and microvascular disease. In the many clinical populations that I've shown you pet derived CFR has been shown to be an independent and incremental marker of cardiovascular risk. And emerging data now suggests that pet derived CFR and coronary flow capacity may be useful in improving our understanding of CAD and its treatment and coronary flow capacity may be a more powerful risk discriminator than CFR alone. So hopefully I've given you guys a few pieces of new information. Before I finish I want to acknowledge a few special people. First my wife Shipra for her unwavering support through the inevitable ups and downs that come with an academic career and for taking care of these two little people while I'm here with you. Sonovie is our four-year-old daughter and Souven's our two-year-old son and no matter what I've done in research or what I do in my career these two will be my greatest achievement. I also want to acknowledge some folks here so again Dr. Willerson and Dr. Coulter for inviting me to give this talk. Dr. Coulter and Dr. Stainback for being phenomenal mentors and making sure that I was well prepared to do the work that I did after I left. Dr. DeCarly for all of the guidance that he gave me for the work that I highlighted for you and all of the other fellows that he's trained. I want to thank Joji and Dew for their warm hospitality. Their kids are awesome and I enjoyed playing with them yesterday and I look forward to tonight at your house as well. The boys and now the men of GCG Steve, Chavik Ali and Angelo and I want to thank Tracy and Kerry for setting everything up for me today. Thank you all so much for being here and making this a really great homecoming. Thank you.