024 - From Risk Factors To Real Disease: How AI Plaque Quantification Is Rewriting Cardiology

Good afternoon. My name is Udo Hoffmann, and I'm very pleased to welcome you to what I think will be an exciting webinar, From Scan to Strategy: Driving Evidence-Based Decisions with Cleerly Compare. I'm the Chief Scientific Officer at Cleerly and also the Chief Medical Officer. My background is I'm a cardiovascular radiologist. I spent about 25 years at Mass General Hospital and Harvard Medical School and have been deeply involved in the development of the field of cardiovascular CT. However, today it is my pleasure and privilege to introduce our 2 speakers, Dr. Alberto Morales and Dr. John Osborne. And I want to briefly introduce Dr. Morales. Dr. Morales graduated from the University of St. Eustatius and did his internship and residency in Georgia at the Regents University. He has board certifications both in cardiovascular medicine and internal medicine, and he's level 3 certified in every advanced cardiovascular imaging modality. Dr. Morales is the owner of South Tampa Cardiology and Advanced Imaging Center. He's spearheading a transformative shift in cardiovascular care by employing cutting-edge imaging technology for the direct detection of heart disease. Recognizing heart disease as the leading cause of death in the United States, where the first symptom is often sudden death or heart attack, Dr. Morales emphasizes the importance of proactive screening protocol before symptoms arise. Dr. Morales advocates moving away from indirect testing such as stress testing instead of using— instead using direct cardiac screening using advanced cardiac CT technology. He believes heart disease should be screened for earlier, much like cancer, rather than waiting for symptoms. Overall, Dr. Morales is a trailblazer, and he has done more than 10,000 cardiac CTs in his practice. And we're very much looking forward to listening to his experience today. Dr. John Osborne obtained his bachelor's degrees with honors at Penn State and his MD with magna cum laude from Jefferson Medical College and his PhD in cardiovascular physiology from Thomas Jefferson University. After graduation, he continued his postdoctoral training at Harvard Medical School and Brigham and Women's Hospital in Boston. Massachusetts, where he completed his internship, residency, and research residency in internal medicine and a fellowship in cardiovascular disease with postdoctoral training in molecular genetics, positional cloning of cardiomyopathies in the Seidman Laboratory in the Department of Genetics at Harvard Medical School. Dr. Osborne published many articles on atherosclerosis, vascular biology, and molecular genetic. His training focused in non-invasive techniques, and he's board certified in cardiology and internal medicine, echocardiography, nuclear cardiology, clinical lipidology, and in cardiovascular computed tomography. So an expert in all associated areas in the treatment of cardiovascular disease. He has been recognized as a fellow of the American College of Cardiology and a fellow of the National Lipid Association. In addition of being the medical director for Market Dynamics, he is the founder and director of State of the Heart Cardiology in Dallas, Texas, and is currently serving as the CEO of Metroplex Cardiology, a preventative and consultative concierge practice focused on early cardiovascular disease detection and prevention using evidence-based approaches and remote patient monitoring. And beyond that, Dr. Osborne is a great human being and a great friend of mine. Before Dr. Morales and Osborne will share their experiences, I will give us a head start here. As I already mentioned, heart disease unfortunately remains the leading cause of death, causing more deaths than all cancers combined. And when you look at the comparison of the state of the art of cancer care on the left and cardiac care on the right, you can see a glaring difference. Cancer care has moved from traditional breast exam, blood testing, or chest X-ray to image-driven detection and screening of disease using mammography, colonoscopy, and lung CT scan. In contrast, cardiac care remains kind of in a state of, you know, indirect detection of signs and symptoms of coronary artery disease. We still base our management mostly on risk factors, although We know that 70% who suffer heart attacks have normal cholesterol. We're waiting for the occurrence of symptoms, although we know that 50% of heart attack victims have no symptoms prior to the first event. And if patients come to us with symptoms, we still do mostly stress testing. And stress testing images, of course, dissectually of coronary artery disease with myocardial perfusion. And 75% of heart attack lesions are missed by stress testing. Cleerly has a mission to change that and to help patients and care providers to see what matters more. And we do this by providing tools that effectively assessing coronary artery disease by characterising and quantifying plaque. Stenosis and ischemia. On the right side, you can see a screenshot from our web-based software. And, uh, what the software does, it provides, um, detailed, uh, characteristics and quantifications of plaque, as you can see here, including plaque composition, stenosis. And it uses a machine learning algorithm including all vessel-based characteristics of coronary atherosclerosis to determine whether the presence of ischemia is likely or unlikely. This helps patients' eligibility for cath pathways like medical therapy or invasive coronary angiography. And it has been designed to be easily understood by patients and by non-expert physicians. As we all know, not all plaque is bad. And knowing about this is very important when you talk about coronary artery disease. Here you can see a graph that shows the various levels of risk for a heart attack that are associated with different types of plaque compositions. On the left you see in blue a calcified plaque that is at very low risk to cause a rupture of a plaque and subsequent heart attack. And you see that when the composition goes towards non-calcified plaque and then even more so towards low-density non-calcified plaque, these— the risk of an individual lesion to cause a heart attack increases dramatically. We also know from IRIS and other data that we can transform non-calcified plaque to calcified plaque by certain medical therapies. So overall, we try to provide the field with a tool that allows for a comprehensive approach to quantificate— for quantification and characterization of clinically actionable findings of CAD, including atherosclerosis volume, composition, length, and location, stenosis both in 2D and 3D, and ischemia end-to-end for determining the likely presence or absence of ischemia. This ends my introduction part. And I now turn it over to Dr. Morales.

Thank you, Dr. Hoffman. Uh, thank you, Cleerly SECT, for inviting me. Welcome, everybody. Looking forward to discussing some cases and how we use Cleerly AI to, uh, better manage our patients and get specific information that otherwise is not achievable through just simple direct CT imaging. So I'd like to start with, you know, what is the risk based on burden? And this was a nice eloquent study looking at 10-year outcomes in patients with the different staging systems, with clearly with stage 1 being percent atheroma of between 0 and 5%, stage 2, 5 to 15%, and greater than 15%. And these patients, as you increase in your atherosclerotic burden, And this is independent of stenosis, Agatston scores, or even ACVD risk scores. You saw that the mortality for stage 3 was well over 15% over the next 10 years and 10 to 15% for stage 2. This is a nice graph showing the different, you know, the different atherosclerotic burden and Hounsfield units, how we measure these plaques. And as you go from a very broad Right. Blue, the lowest risk of calcific disease, all the way down to the lowest Hounsfield units of less than 30. You could see the risk incrementally increases over this lipid necrotic core of -30 is that very inflamed plaque that we suspect is intraplaque hemorrhaging right before an event. And this makes it the most Highest of risk. Typically, minus 60 has been used traditionally in other studies as as the lipid-rich necrotic core. However, minus 30 is a much more specific and more more significant terminology that we'll be reviewing here. This is a nice graph I put in, basically looking at all the studies out there, looking at all the plaque burden, and and based on risk, what is your likelihood of, you know, if someone's diabetic versus non-diabetic, what is their yearly progression? And this is measured by total plaque volume. And as you can see here down at the bottom, diabetes on a— if you're a diabetic and you're found to have stage 1 disease at the time of diagnosis, let's say in your 40s, and you do nothing different, this patient's gonna increase at a total plaque volume of about 50 cubic millimeters per year.

Wow.

And if you add incremental risks such as high LDL or elevated BMI, you know, this progression rapidly increases over time. And there's data that basically discusses that if you're found to have stage 1 CAD, that over a period of 10 to 15 years, you would develop advanced stage 3 disease. And so you could see how if you can optimize these risk factors by lowering LDL, lowering ApoB, how you can start inducing regression and also with certain therapies that we'll discuss. So let's talk about some cases. First case is a 65-year-old female we saw in our clinic. She was referred to us basically because she had an elevated calcium score. She was very quite shocked to this. She was asymptomatic. She was a vegetarian all her life. Really, her only risk factor was that she had rheumatoid arthritis, and she was basically stable on medical therapy with— on biologics. And as you can see here on her CTA, she had significant burden of disease. The LAD on the left-hand side had, you know, diffuse moderate disease extending from prox to mid. The proximal RCA also had somewhere between 50 to 69%. And on the little finger pointing there, you can see the left circ and obtuse marginal branch had this little tiny plaque, less than 15, 25%. This overhead view showing the extent of the disease extending into the LAD and the RCA, and then just very minimal disease extending into the left circ OM system. So this is a nice view of Cleerly's, you know, percent stenosis. And when we add And as you look at the plaque burden, you can see there's extensive plaque burden in the RCA and the LAD system. You can visually see it, but it's very hard. And so if you add the Cleerly, there you go. You can see that the soft plaque is a little bit much easier to recognize. That soft plaque, and I'll go back, that soft plaque in that RCA is a little difficult to see. between the epicardial fat and those layers. So the AI does a really nice job of enhancing that. And then as you can see, this volume is able to be measured. And so you can measure this entire volume of all these plaques over time in the coronary system. And we call that total plaque volume. You can also measure the percentage of that plaque filling each of these arteries. And we call that percent atheroma. And then, in addition to that, you can actually characterize these plaques based on morphology, whether it's soft plaque being yellow. If it's blue, it's calcified plaque. And then these little red dots in this RCA are very warning signs of low attenuation plaque and plaque that's very vulnerable and ready to possibly rupture over the next 6 months to a year. As you can see, There was no significant stenosis. All the lesions were between just around 50%. Then Cleerly has a very nice ctFFR. It's binary, uses machine learning algorithms that have been validated and correlated with invasive FFR and several key landmark trials—

Excuse me.

—that have, with a very good area under the curve that's similar to the other platform, ctFFR, HeartFlow, correlated very nicely, in some cases actually better than HeartFlow. And as you can see in this patient, no significant burden of ischemia, maybe some mild ischemia at the bottom there in the PDA territory. Now, what I would say is that most of us would stop right here. We would say, hey, this is someone with 50% or less stenosis. No significant ischemia, and we would, you know, basically manage this patient. Some of us might even just do a stress test, which would be negative, and we would say, you know, this patient's risk of having an event is low, or low at moderate level, and we would treat this patient, you know, conservatively. What I would argue is what I have already shown you is significant high-risk features, and let's go over that. So, You know, we know that a lot of patients who have acute coronary events, roughly 70-75% of patients with heart attacks have non-obstructive, less than 40% stenosis. We also know that a lot of patients who have heart attacks are asymptomatic, up to 50%. 25%— 25 to 30% of these patients who have heart attacks have zero calcium scores. So there has to be more information out there to be achieved to label these patients. Why are they having events if they don't have significant stenosis? I think we need to move beyond stenosis and beyond ischemia. What I like to— what I usually tell my patients is that modern cardiology isn't about, is there blockage, is there ischemia, and if there's not, we're done. It's about how much plaque volume is there, how much plaque is there, how much, what type of plaque there is, and what type of plaque is, and the risk associated with that plaque. And then can we track this plaque over time and monitor it? So this is a nice, well-studied, you know, comparison of all these plaque features. And I'm just gonna, you know, there's a lot of metrics out there, but we're gonna review these 4, which is total plaque volume. Anything greater than 564 cubic millimeters is basically a hazard ratio of 2 to 2.5 above and beyond high percent stenosis. So as you can see here, greater than 70% stenosis has a 1.5 to 2.0 ratio. What that means is basically patients are either at 50% or 2 times likely to have a heart attack compared to patients with no stenosis. When you add this extra information, total plaque volume, percent atheroma, you could see how the risk is magnified and twofold on top of percent stenosis, a higher risk morphology based on this plaque information. So there's plaque information to be obtained, and you need this Cleerly AI or any AI company to kind of analyze it, and Cleerly does a superb job in this This is how I use my information. So this is the full breadth of Cleerly's atherosclerotic stenosis report. And what I would say is that we need to go above and beyond looking at just percent stenosis and ischemia. And I know that this information can be sort of overwhelming with all these numbers, but as you can look in this patient, 846 is a significant total plaque volume.

Uh-huh.

Goes above and beyond risk. This low-density 18.4 is significantly high risk for an event within the next 6 to 12 months. Non-calcified plaque volume greater than 200 is associated with a threefold risk in heart attacks. And then this percent atheroma greater than 15% is associated with a greater than 15%. This signifies the plaque burden in the arteries. So let's look at this patient. So she came to us with these labs. Her LDL was in the 133 range. She had all these, you know, all these markers, these 4 metrics, significant total plaque volume, high-risk plaque features with low density. We offered her aggressive medical therapy, even antiplatelet therapy. She declined all these. And so we didn't hear from her for another 2 years. And she comes back after a hospitalization 6 months prior, and she was hospitalized for pneumonia, and she had a bump in her troponins, and cardiology was consulted in-house and had deferred any further testing because of her infection and suggested she go reach her cardiologist as an outpatient. So we saw her, and she was having this progressive shortness of breath on follow-up, and we decided to rescan her. And as you can see here, She's had significant progression of disease on her follow-up scan. Actually, these 2 are flip-flopped, but the middle image is the LAD, and you can see extensive progression in her LAD, significant progression in her RCA. And then on the right, you can see where that, you know, paradoxically, you saw a very mild plaque previous to that, and you can see complete occlusion of her left circ OM system. Here's a sort of a before and after from 2 years ago, and you can see she had a very non-obstructive looking plaque that eventually became inflamed and ruptured and caused her acute coronary syndrome. And then above you can see a significant change in her anatomy and obstruction of disease. So this patient, obviously she had all these high, you know, inflam— you know, all these high metrics that put her at significant risk. And so this became more of an observational evaluation of a patient, and she had obviously significant burden of disease. And this is her follow-up scan, and you can see her— not only her disease burden has increased, her percent stenosis has increased with a new occlusive lesion. You can see that the low attenuation, these little red dots, have progressed in size. So let's look at her. And as we look at her ischemia, you can see she's had progression in ischemia. So at this point, she's progressed in just over 2 years from a more advanced picture. So clearly does a really nice job. There's this comparison feature where you can compare one scan to the next scan, and you can assess it by cubic millimeters, by percent atheroma, you can Compare it by volume or even by percentage, and you can track disease over time. And as you can see here, she— this patient had a cubic millimeters total plaque volume of 846, and it had increased by almost 140. And this doesn't even take into account the complete occlusion. So if you include that, it's significantly more. The non-calcified plaque has continued to progress. at a rate of almost 70 to 80 per year. So she's continuing to build this low-density, this non-calcified, very vulnerable-looking plaque. Underneath that, you can see the low attenuation plaque. Any plaque greater than 2.4 puts you at fourfold the risk of having an event over the next 6 to 12 months. And you can see she continues to progress in this fashion. And then the calcific disease did not significantly increase. We consider this a nice change if it increases and you decrease your total plaque volume. Here's a nice summary view of this change going from, as far as percent change, increase in total plaque volume, which implies increase in plaque burden. The increase in non-calcified is more of this active plaque that is very risky. And then obviously almost a 50% increase in low attenuation plaque. Before I turn it over to Dr. Osborne, I'd like to point out 2 obvious things in this case. One's obvious that this patient, you know, is a prime example of a lot of markers, 4 markers that are significantly elevated, leading to a significant risk in this patient. And then number 2, It wasn't the suspected high aggressive-looking lesions that ruptured. It was actually the non-obstructive lesion that ruptured. And with that, I'll turn it over to Dr. Osborne.

Thank you, Dr. Morales. That was an amazing natural history example of cardiovascular disease. Fascinating, scary, and tragic all at the same time. So what a beautiful illustration. And you also beautifully teed up all the science and kind of what all these numbers mean. So now we can have more fun that everybody is on the same page, kind of going through these compare cases and what happens to plaque over time. I think the other message here, and what I'm going to hope to show, is that at the end of the day, we really want to make sure that historically we've used numbers, right? Blood pressure is good, hemoglobin A1C, is good, cholesterol's good, ApoB, et cetera. But I think with the ability to use AI and clearly, and now clearly compare, we can actually measure the disease. And I think that ultimately needs to be the ultimate marker of success, not whatever LDL is or ApoB or Lp, et cetera, but what is, what's happening with the plaque. And I think we've got some great cases of mine and Dr. Morales that'll really illustrate that beautifully. Here we have a 54-year young Male, had some atypical chest pain a couple years ago, had a family history of early premature heart disease, never used tobacco. He's ripped, works out 6 times a week, aerobic and isometric, normal Lp, was on no meds, blood pressure reasonably okay for a 54-year male. You might question that. Heart rate, BMI is 23. These are his incoming lipids, by the way, on no medications. So, no diabetes, GFR, LFT is okay, blood sugar was fine. No LDL 56, triglycerides 48, HDL 51. And I think, as I recall, yeah, he had actually had a calcium score, I think, before he saw me, and that was 0, for the record. So, because the family history, I know his numbers look great, but he had that family history. Some atypical chest pain. We said, let's do cardiac CT. So I'm gonna show you with and without the AI just to illustrate how impressive the AI, both impressive and humbling. So I've done about 25,000 cardiac CTs over the last 25 years. And I would say, looking at these images, and I know that this is hard to see, but I called this a normal study. I just didn't see any plaque. Well, let's turn on the AI. When we turn on the AI, there it all comes out. So lots of low-density lipid-rich plaque here, all the yellow stuff in this vessel is his right coronary artery. So we— I'll show you in a minute his net numbers because we did do a compare. So at this point, even though his lipids look great, we had put on rosuvastatin and some olmesartan to optimally treat the blood pressure and and lipids. He comes back 2 years later, again, no symptoms, feeling great and fine. So, here's what happened in 2 years. We saw nearly a 50% reduction in plaque. Remember, he had no calcified plaque initially. He had a 50% reduction in non-calcified plaque, a 70% reduction in the always ominous low-density non-calcified plaque, And here, interestingly, as we would fully predict from our biology and understanding of this disease and our understanding of the pathophysiology, that now, after a couple years of treatment, getting his lipids even lower, he now shows up with a tiny little itty-bitty speck of calcium as well. I think too, this was fascinating. Several messages. One, never trust a zero calcium score. Number 2, labs may not tell you what a person's risk is, certainly as we looked at his LDL labs and even Lp. And also, too, that for me, it was striking that he had so much plaque. It was all lipid-rich plaque, really in the wall. As I tell my patients, I talk to my patients about the donut and the donut hole. All of his plaque was in the donut, none of it in the donut hole. Looked to me to be normal. AI really saved me from that. And just illustrating what we can do. And I think too, the other message, this is an N-of-1 experience here in this particular case, but fascinating to see such a striking reduction in lipid-rich plaque. I call that the lava, the calcium is the hard plaque, the extinct volcanoes. But really dramatic reduction in just a couple of years with some modest medical therapy that suggests to me maybe the earlier we treat, the bigger bang for the therapeutic buck as far as plaque goes as well. And this is just some of the illustrations of what happened here. There's the compare that we did that I showed you before. Again, the vessels really look great. It's really the AI that really Helped me to zoom in on the real disease. And I will say one thing before I pass it over to Dr. Morales. I love to describe cardiac CT with AI now as cardiac CT coupled with AI is the cardiac colonoscopy. Let's find the polyps, take care of the polyps, maybe see polyps we couldn't see by eye. But it's not just cardiac colonoscopy. Adding AI is cardiac colonoscopy with histology. So with that, I will pass it over to my colleague, Dr. Morales.

Thank you, Dr. Osborne. Beautiful analogy. I couldn't agree much more than screening these patients very early and the better, much improved outcomes you get when you start treating the disease at an earlier stage. Here's a gentleman who was 57 years old. He was referred to me by his endocrinologist for some prediabetes. And had some, some risk factors, some, some anxiety, reflux, elevated BMI, and really some mild shortness of breath. So we decided to do a CCTA on him, and you could see this high-grade-looking lesion in the ostium of the LAD. A mixed plaque. I called it 70%. And the, the catch here is that this patient had some abnormal labs, had very low platelets around 50 to 60 range, had some low white count and some low hemoglobin. So some pancytopenia, and there was some, obviously some undiagnosed condition. He subsequently did end up getting diagnosed with MDS, but this added to the conundrum here of this patient. So looking at the Cleerly AI, and you can see, you know, there's a significant burden of plaque extending from the left main into the LAD. And when you add the Cleerly AI, you can see it light up like a Christmas tree. And you can see most of this plaque is bulky, it's soft, it's non-calcified. You can see some calcified components as blue and the non-calcified components as soft. You can measure this plaque and this plaque volume was extensive, especially in the proximal AD. He did have some disease in other segments, but most of it was— loculated here. Clearly, I called it 66%. And as we extend into the ischemia platform, you could see it was labeled significant ischemia, all extending down into the diagonal system. Again, most of us would stop here, send them to the cath lab, he'd probably get a stent. In our case, we saw some numbers that were quite alarming. And we, you know, obviously the concern is putting a a fresh stent in a gentleman with a bone marrow problem who may require a bone marrow biopsy, chemotherapy. And so we discussed options with him, and the end result was that we would manage this patient very aggressively with medical therapy. The problem lied at what medical therapy? So let's get into— this is his clearly AI atherosclerotic and stenotic burden. And as you can see here, You know, a significant risky-looking lesion in the prox LAD with ischemia. He has a significant amount of total plaque volume at 763.9. If you look in the middle, low density, there was no significant low-density component. Again, this is, this is where I tend to gravitate to first. I look at this first. This tells me what their impending 6 to 12 month risk of having an event. If these numbers are well over 2.4, I get very concerned and I want to treat this patient very aggressively. The next thing I look at is their non-calcified plaque volume. Anything greater than 200 puts them at threefold risk of having an event over the next 1 to 2 years. And so, so, so this is a significant bulky amount of plaque. And then, you know, obviously looking at total plaque volume and percent atheroma and tells me how much disease and how diseased are they in their arteries and what is their 10-year risk. In this patient, very high at over 15%. Then here's the AI analyzing it and giving you the breakdown of the plaque characteristics and volume. Let's look at this case thoroughly. When I saw this gentleman, he weighed 220 pounds, His primary care already, he had already put him on Crestor. He'd been on Crestor for years and had been increased to 40 a couple years prior to him being referred to me. And if you look at his numbers down below, he already had an LDL of 32. His Lp was the only thing we hadn't checked, and his A1c was 5.3. As you look at the follow-up labs, you know, he was, his Lp was -10, so it wasn't playing a significant role. You know, we tried to get him on PCSK9. Really, there was little, little reason to put him on that. And so the only thing we could offer this gentleman was, you know, weight reduction. And, and, and, and we started him on GLP-1 Ozempic, uh, with its benefits of 20% reduction. We, we, we said, let's, let's get you on something that's going to reduce your events, help you lose weight. And, uh, he did end up having about a 22% 22-pound weight loss over a period of a year. And so he was doing well throughout all these regular follow-ups. And on the 1-year follow-up scan, we rescanned him and you could see a significant improvement in his prox LAD. We called this less than 50% partially calcified plaque. And just an up-close view from 2023 to 2024, you can see a bulky— and the majority of the improvement again lies in the non-calcified component. And I'll share with you what that data looks like. And you can see by clearly there was a significant improvement in percent stenosis reduction from 66% down to 47%. And as we add the analytics to it, you can see a significant change from that bulky soft plaque being kind of shrunken down and a shift from soft to blue. From yellow to blue, implying this healing and stabilizing of this plaque and improvement in percent stenosis. And as you follow it up with the ischemia, you can see a significant reduction in this overall burden of ischemia, not completely resolved, but definitely improved. And at this point, he was, you know, had lost 22 pounds. He was feeling great, had no further symptoms, and was feeling great. So let's take a look at the— just, just a visual assessment. And on the left side of the screen, the first 2 you could see without the clearly color coding, and on the right you can see a clearly color coding. And then on the bottom you could see the change from 2023 to 2025, and you could see a significant improvement in reduction and a shifting from non-calcific to calcific. Now you can probably visually call this But you need AI to kind of help you quantify this. The AI is just like Dr. Osborne said, the histology of a CT scan. And so, and then the comparison is where you're gonna get to see a lot of this information. So let's go to the next slide here. So you can see the total plaque volume went from 763 to 752. So we reduced— 11.4 cubic millimeters of plaque. That's a 1.5. So this is in a non-diabetic, you typically get 21 to 20, and then elevated BMI, you get an additional 15%. So this gentleman should have probably increased by about 30, 35 cubic millimeters of plaque per year. And in 1 year, we subtracted about 11.4. So this is slight. This is a true case of regression. And the non-calcified volume, you can see, went from 579 down to 383. This is almost a 200 cubic millimeter drop, almost a 34% drop in this plaque. This is the active plaque, the plaque that shifts, that could turn into this low attenuation plaque. And again, had no significant low attenuation plaque, so no significant change there. And the benefit is you can see his calcific plaque went from 184 to 369. This is almost a doubling effect of the plaque volume. And this implies plaque healing and stabilization with medical therapy. And again, this is a nice case where Ozempic was the culprit here because the patient already was on a nice dose of a statin. And the only thing added was— GLP-1 with significant improvement in weight reduction. So 2 points on this particular case. Obviously, we know the benefits of GLP-1s and 20% reduction in MACE events, but what is that due to? I think there is potentially a direct anti-atherosclerotic effect with these GLP-1 inhibitors. And then, and number 2, obviously, You know, numbers, you know, this gentleman had an LDL of 33. You would have thought that these numbers were protective and he shouldn't have significant ischemia or significant disease. And obviously, if you only look at these surrogate markers and you don't look at the disease, I think you're going to miss a lot of potentially risky patients, risky plaque, and events. And so, Being able to analyze with CT but using AI is— and then this Clearly Compare, which allows you to track disease over time, allows you to track the therapies that we're placing on these patients to have improved outcomes. With that, I'll turn it over to Dr. Osborne.

Wow. Thank you, Dr. Morales. Fantastic cases. I really have to hand it to you. Just beautifully Done very illustrative and a beautiful illustration of what medical— good medical therapy, utilizing all of the evidence now, of course, the GLP-1s now with the SELECT trial, also illustrating the cardiovascular event reduction, and just a beautiful illustration of that. I will say, I love to coin terms. I've got another one for you. I've not been in the cath lab for 20 years because of cardiac CT. I was never an interventional guy. But now with cardiac CT, AI, and all of the evidence-based incredible array of tools with multiple different mechanisms of actions we can use to treat plaque, to stop plaque, to reverse plaque, I like to call myself an interventional endocrinologist. So apologies to any endocrinologists out there, but I think it's— that really describes well what we're doing with atherosclerosis in the 21st century. And I really do believe, as you had said too, or paraphrase you, that cardiac CT with AI is the essential primary tool for evaluating patients with or at risk of cardiovascular disease and now being able to track it. In fact, when you think of it, until we had cardiac CT with AI, not just cardiac CT, we really had no ability to say to a person, look them in the eye, and say to that person, your plaque is better, or your plaque is worse, maybe in a semi-qualitative way, but not with a quantitative way that we have with cardiac CT and AI and quantitative plaque analysis with Cleerly. So another case, I think this is another case, 62-year-young physician, asymptomatic, had a very high calcium score. I forget how high, but—

1,000.

probably 2,000 or so a few years ago. Hypertension, hyperlipidemia, was significantly overweight as well, but had hypertension, hyperlipidemia for 11 years. Normal Lp, never used tobacco. He's on quite the litany of kind of maxed out drugs, maxed out statins, Repatha, aspirin, olmesartan for his blood pressure, tirzepatide, lost 50 pounds. Blood pressure really Improved dramatically, but he's still on some omeprazole for that. You can see his vital signs. Everything looks good. He now has a BMI of 24. When we did his initial scan, there's his initial numbers. All the numbers look excellent as well. This was his initial scan. So we did the initial scan in '23. He had quite a high plaque volume, so 1,200 over 1,200 total plaque volume. A lot of it was non-calcified plaque. Fortunately, no low-density plaque, and of course, the calcified plaque. So, at this point, he got the full press. So, we added ezetimibe, we added colchicine, we thought about rivaroxaban and aspirin, or a P2Y2 inhibitor, and the plan was to rescan this fellow here. So, we did so. And so this is initial scan. This is the follow-up scan. And I will mention and reference a very, I think, pretty great paper. I'm on that paper, but with a variety of other preventative cardiologists published in the American Journal of Medicine. We published, I think, January of '23, that talks about the staging of this disease. So in this case, he would be a stage 3 patient, which is the highest. And based on the staging at both suggest different therapeutic modalities based on stage, but also when we should rescan him. So if you're stage 3, which is above 750 cubic millimeters of plaque, the suggestion is that we should rescan those patients in a year to make sure that all of the modalities that we're throwing at this particular patient are actually working. So we Did all that stuff and we rescanned him. And here again is— this will be an open question case because his scan is scheduled. It's going to happen this July. But despite all of those great numbers, all the right anti-atherosclerotic therapies, all evidence-based, his plaque still progressed. So total plaque went up, Not terrible, but more concerning is his non-calcified plaque really went up. He had an increase in his calcified plaque as well. So, this one I'm going to string you out on a little bit because all of these things we did after this scan in '24, and our plan is to scan him in '25. So, this is also a hook to keep you guys coming back for the next time, and I'll show you what happened after we put the full-court press with additional lipid therapies and some anti-inflammatory therapies as well. So I'm going to pass this back to Udo. Okay.

Thank you, Dr. Moraes and Osborne. I mean, it's great. We're making the tools. We're trying to provide the tools to manage patients based on atherosclerosis, which we think is the right approach to do. And nothing beats to listen to 2 practitioners who actually use this enthusiastically to treat their patients and see the success stories. And, you know, in case of the natural history, you know, case, also what happens if you don't do it. So it's really amazing. Thanks to both of you. This is a conclusion. I would say, you know, the conclusion for me is, you know, I think treating patients based on atherosclerosis is the right thing to do. You need an accurate tool that quantifies plaque at baseline, both for composition, volume, but then also is able to tell you the changes over time. And together with the stenosis and ischemia assessment, this really raises confidence in how you manage your patients. You know, you, we have seen cases where, you know, plaque imaging really goes, allows managing patients and treating patients beyond their LDL values. And I just want to, you know, you had a couple of cases that were excellent.

Yeah.

even if your LDL is in a normal, you know, in the normal range, you know, consider or you need to be aware of that. It doesn't matter when you lower your LDL, the results that we have seen in trials where every 39 millimoles per deciliter you reduce the cardiovascular event rate by more than 20%. That holds true even for when you reduce an LDL from 50 to 30 to 20. 4 to 10. So, you know, that, that I think is fantastic, needs to be considered. And I, I do have a few questions, you know, to— for both of you about tracking autosclerosis. But you have shown, you know, that it's highly valuable to monitor the responses. And with that, I think So please, we have already a couple of questions, but please feel free to add more questions in the Q&A. In the meantime, I just wanna, you know, tell you what's next for CLEARLI and hope, you know, that, that some of you may be interested to, you know, learn more about CLEARLI after today's lecture. We, we have a workshop in Louisville, a learning lab workshop with live CT cases in Louisville, Kentucky, July 9th. And then at the SCT the week after in— We have a hands-on AI-powered CT workshop from 5 to 8 with dinner, with a real dinner and drinks. So, you know, feel— please feel free to register. You'll see the link here. And then, you know, as we should be, we are very active with the American Society of Preventive Cardiology, and we will host a dinner symposium there with some fantastic speakers. Okay. So we have about 10 minutes. So, you know, before we— we have 8 questions. Before we go there, May I ask Dr. Osborne and Morales, may I ask you, the, you know, how do patients react when they see atherosclerosis, you know, both at baseline and then also, you know, if you present to them the follow-up results? I mean, have you seen a difference as opposed to when you talk to them about LDL values?

I can start. So yeah, I think the power of show and tell is very powerful. especially when you're trying to convince patients to take certain key medications. When we do these scans routinely, I typically take image shots of all their coronary images. And in each patient exam room, we have examples of normal versus mild versus progressive disease. And when they see their images and they see their Cleerly AI reports and all that quantitative information, it can be overwhelming. But The power of seeing their disease and comparing it enhances your credibility as a physician to empower them to do something about their disease. And then the ability to tell them that we are going to do A, B, and C, and then we're going to follow up by taking another set of images to see how this therapy is working adds this this sort of taking it back to the patient, right? I mean, when most patients come to the doctor and they have a problem like high blood pressure, for example, you know, the doctor's gonna measure the blood pressure, he's gonna give 'em a value, right? And then we're going to put 'em on a round of blood pressure regimen, and then when they come back, we're gonna recheck their blood pressure. Well, we fail to— first, we don't really check for CAD. We look at all these surrogate markers, But we have a way to check it. We have a way to measure it. We have now very good therapies out there, and now we have a way to measure success and whether or not that therapy is translating to success. Oftentimes it does, sometimes it doesn't, as we've seen with Dr. Osborne and several other cases where patients have been on very good effective therapy. The numbers look good, but yet the disease doesn't. And so it allows us an extra layer of therapies to try to improve and change the outcomes.

Thank you. And, you know, there have been studies many years ago, even with coronary calcium, and even recently out of Stanford that showing, I mean, most of these therapies are lifelong therapies, right? And now, as you mentioned, we have so many effective therapies. It's not that we lack therapies, but our biggest hurdle is getting the patient to take the medication. And so, you know, many or several studies have shown that compliance with medical therapy increases dramatically once they actually see their disease. And so I'm very glad that your experience is similar. And John, I know you wanted to say something.

I was going to say, I think Dr. Morales is absolutely, as usual, right on target. And what we do is I'm going to call it radical clinical transparency. So every patient sees his or her images. And when we do the compare, so that, in fact, by the time we do the compare, our goal is to combine the radical clinical transparency with education so that when they see the images, or, and even better yet, the compare images down the road, we both go— myself and the patient both go, yeah, this is great, or we got other things to do. And it's amazing to think that we practice cardiology, cardiovascular disease, most of which is oriented around atherosclerosis, for well over 40 years. And we never had the ability to say to the patient, your disease is worse, the drugs aren't working, or disease is better, the drugs are working, or we need to change our strategy. We could never do that until we had CTA coupled with the quantitative plaque analysis with Cleerly. It's a breakthrough technology. So thank you, Udo, and everybody at Cleerly for giving us this amazing tool.

Absolutely. Thank you, John. Hey, let's go to a couple of these questions. Okay? So the first, or more like, what is the reference for the slide that showed regression with LDL less than 30? I think, Dr. Marles, that was your slide, but I'm not quite sure.

What was the question again?

Oh, okay. Yeah, it was just what is reference for the slide that showed regression with LDL less than 30?

The GLAGOV trial, I suspect.

Okay.

Yes, yes, there's a couple there. There was, there's, I think, 4 or 5 studies that looked at that with intense therapy, lowering lipid, lowering LDL less than 30. There was significant regression. A lot of these studies were IVUS studies, but there are some CT studies out there as well.

Yeah.

Yeah, in fact, you know, in the GLAGOV trial, if you were randomized to the PCSK9 added onto your statin with an LDL achieved about 30, 83% of those patients showed plaque regression with IVUS. I like CT better for any number of reasons, but yeah.

I think it's really important, you know, it doesn't matter whether your LDL is in the normal range, it is the same effect if you lower LDL even from there as if it's higher, so that effect is maintained. Here's a many study has used less than 30 harns per units for low density plaque. That's correct. I don't know what that is less than 30. That seems to be the same. I think low density is safe to say is less than 30 and is above fat. So it's above minus 100. Can we reverse hard plaque? Which is What do you, what do you, Sanam, what do you think? Hard plaque, calcified plaque?

I think it's very difficult, although Dr. Budoff just published a recent study that looked at colchicine. It may have some benefit in regressing calcified, heavily calcified disease, but there isn't a whole lot published on that. Most of the literature is in non-calcified plaque and mixed plaque.

Here's another question. How did Cleerly validate their FFR-CT software? How does it compare to hot flow FFR-CT? So I can take that one. So we, we have extensively validated this against invasive FFR, which is the gold standard. But moreover, we now also have validated it against O15 water pad, so myocardial blood flow. And so, you know, ischemia positive is correlating to a decreased myocardial blood flow as measured with O-15 water PET, which, you know, which is the best measurement. Unfortunately, it's not available in the US yet, but there is, there's an active trial to potentially get an FDA approval. And then how does this compare to heart flow? Again, you know, we are happy to send you the publications if you, if you like. I think there's an email, Lauren, that we can send it to, or we can, if you, you know, we are happy to send you the publications. I can say in general it compares favorably to hot flow FFR CT when compared to invasive FFR. Next question here. 2 minutes. How much can we lower LDL and still be safe?

I think, you know, LDL is a very good marker to start with, but, you know, I think, you know, when you're looking at the spectrum, there's other markers such as Lp and ApoB, which gives you really all the atherogenic particles And so we've really hampered more ApoB reductions, and we're trying to shoot for under 50 for ApoB for plaque regression, and obviously for plaque stabilization and stability, less than 80. And, you know, with Lp, we're seeing, you know, there's been some recent studies that look at independent of how low your LDL, you continue to progress in non-calcific, especially low attenuation plaque. in these patients with high Lp. So obviously there's some treatments out there in the pipes that lower it by 80%. Right now the best therapies are drugs like Leqvio and PCSK9 therapy that can at best reduce it 30%, maybe in some cases up to 50%, I've seen in some of my patients.

Yeah, thank you. And I think here's a great message. So me as the lipidologist, we've always used our lipid profile, LDL, It's useful because we want to know, do we need to target LDL or triglycerides? So that's useful. So the lipid panel still has value. But really, really, I'm going to encourage everybody out there, get an ApoB every time you check a lipid panel, because that's really the, the, the main, uh, the best, uh, uh, marker of lipid risk, uh, from lots of studies. And what you'll find is that when we get LDLs to low levels, Even if you use Seth Martin correction, etc., it's just not accurate. So use ApoB and you'll see, you know, I'll end up with patients with LDLs of zero, even negative, right, with the old Friederichsen model. But when you use ApoB, yeah, it'll definitely come down, but it's nowhere near levels that we would consider pathologic. You know, you get ApoB to 4 or 5, that's what we see in hypo and A-beta lipoproteinemia. That could be an issue. I've never seen that happen even with people on 3 and 4 LDL-lowering drugs. So it's essentially impossible to make— to suppress ApoB LDL to levels that would be harmful.

Yeah, thank you. You know, the time flies and we're over time. The hour went by very quickly, but thank you again, Dr. Osborne, Dr. Morales, and also to the audience for the good questions. And I hope we all learned We all learned something useful today.

Thank you very much. Have a great day.

See you at SCCT and ASPC.

Exactly. Bye-bye.