 Welcome to Texas Heart Institute Educational Programs Cardiology in the Time of COVID-19. In response to the evolving COVID-19 pandemic, the Texas Heart Institute is providing perspective to the healthcare community taking care of cardiovascular patients with COVID-19. The title of today's program is the role of cardiac CT and FFR in cardiac patients during COVID-19 pandemic. This program is presented together between Texas Heart Institute and ISCVS or International Society of Endovascular Specialists. I'm Zvon Bue Kraser. I'm an Interventional Cardiologist at Texas Heart Institute and CHI Baylor St. Luke's Medical Center. Our special guest today is Dr. Christopher Zarens. He's a founder and senior VP for Medical Affairs at Hardflow Incorporated. Welcome Dr. Zarens to Texas Heart Institute Educational Programs. Thank you. Pleasure to be here. It's a special pleasure to have Dr. Zarens here with us for this presentation. As COVID-19 situation continues to evolve, guidance documents around elective procedures are being published daily by various entities, government bodies, professional societies, publications and so on. Numerous societies, as I have mentioned, have made recommendations and guidelines regarding treatment of patients with cardiac disease during COVID-19 epidemic. It was shown from several publications that some patients are presenting with elevated troponins and other signs of STEMI or non-STEMI or acute coronary syndromes, and they may actually have COVID-19 related myocarditis rather than coronary event as shown in this particular publication. So it's important to know that myocarditis related to COVID-19 may complicate the diagnosis of STEMI and acute coronary symptoms. Now the advisory board from American College of Cardiology related to COVID-19 clinical implications for cardiovascular patients showed that cardiovascular disease is a significant comorbidity for COVID-19 in addition to age and other comorbid conditions as shown in this particular slide. Also, ACC and Society of Cardiac and Geography in Intervention have made recommendations to avoid elective invasive procedures whenever possible during COVID pandemic. So Dr. Zarens, I would like to know what is your opinion? What is the role of cardiac CT, angio, and FFR in the variation of patients with coronary artery disease in general? Well, thank you Dr. Cratcher for inviting me to participate in this very important discussion, and I wanted to just start by thanking all of the doctors, nurses, and healthcare workers who are working so hard in this very difficult and challenging time to provide care for patients with coronary artery, coronary virus disease. So the problem of diagnosis of these patients is particularly important, especially at a time when many patients are not going to doctors, not going to hospitals, but yet are presenting with coronary disease, which as Dr. Cratcher indicated sometimes is indistinguishable from the symptoms of the COVID-19 virus itself. So in thinking about coronary CT, first of all, we have to recognize how good a test it is. It is the best diagnostic test there is to find out whether or not you have coronary artery disease. It clearly shows you the coronary arteries themselves, shows you the lumen caliber, shows you plaque, and the complications of plaques such as calcifications and plaque abnormalities. And it also tells you about long-term prognosis as shown on this slide. Importantly, patients who have no significant coronary disease on cardiac CT have an excellent long-term prognosis. Some people say 8 to 10 years warranty, and people with non-obstructive coronary disease, the cadrats level 1 and 2 as shown in green, also have a good prognosis. The patients who have problems are those with obstructive coronary artery disease, and the extent and severity of disease is directly linked to adverse events. And in the short term, which is what we're considering now in the coronavirus era, what do we do with these patients? Well, normally when a patient presents with symptomatic coronary disease and has a positive significant obstructive disease CT scan, they're sent to the cath lab. But that, of course, is a difficult decision in a time when we're trying to contain the risk, both for the patients and for healthcare workers, and to reduce invasive procedures that may not be needed in the hospitals. So how do we do this? So let's look at the one thing that we need to remember as we'll see on the next slide. It's important to remember that when we're talking about CT stenosis, there's a disparity between the anatomic stenosis that we see on the CT scan and the functional significance of that stenosis. And the functional significance of coronary lesions can be defined in the cath lab by measuring fractional flow reserve. And we know that visual interpretation of coronary angiograms do not correlate well with the estimates of stenosis. In the FAME trial, patients with 50 to 70 percent lumen stenosis, 65 percent of those patients had negative FFRs that is not functionally significant lesions. Even in patients with 70 to 90 percent stenosis, 20 percent had non-functional disease. So therefore it's important to know about the functional significance of the disease. And we can see in coronary CT the same relationships exist. On the left side of the screen, we can see that for 30 to 50 percent stenosis, one out of three patients will not have a functionally significant lesion. In the 50 to 70 percent range, one out of two patients will not have a functionally significant lesion. So this is an intermediate stenosis group that perhaps really needs special attention. Even in a group with 70 to 90 percent stenosis, one out of four patients will have normal FFRs and not have functionally significant disease. So Chris, of course CT for the relation of coronary disease has existed for a long period of time. And for a very long period of time, it was considered not to be a very sensitive tool. So really what has changed dramatically, the sensitivity specificity and usefulness of CT is the introduction of FFR. Can you tell us how did this evolve and what were the algorithms used to make this test so sensitive and so reliable? So coronary CT scans are actually very sensitive in finding disease. They have a very high negative predictive value and very high sensitivity. They're not very specific, so that a lot of patients who are found to have disease that looks significant on CT actually have non-functionally significant disease. And we have been looking at the relationship between hemodynamics and alpha sources for many, many years. For over a 25-year period at Stanford University, we were looking at coronary, at blood flow and arteries based on human CT scans. These really started with aortas and the problem of aortic aneurysms and stent grafts because in treating aneurysms and stent grafts, CT scan is the workhorse for diagnosis and planning treatments. In 2005, when coronary CT came available with 64-slice scanners, we for the first time were able to see the coronary arteries, not just aortas and carotids. We then started doing computational analysis of blood flow in the coronary arteries. And when the FAME studies came out proving that FFR guided revascularization was better than visual guided revascularization, we then focused on coronary artery flow modeling. And applying the principles of computational fluid dynamics, we could then compute fractional flow reserve. Now as you know, fractional flow reserve is measured in the cath lab by giving a dent scene with maximum coronary blood flow drug-induced, a dent scene-induced. In computational analysis, we simulate the maximum coronary blood flow by reducing peripheral resistance. That then allows us to calculate the pressure gradient across lesions under simulated maximum flow conditions, i.e., computed fractional flow reserve. The problem of all our calculations was we really didn't know if it worked. So in 2009, we actually began comparing our computed values to measured values in the cath lab with fractional flow reserve. And many, many studies now have proved the accuracy of computed analysis so that a computed fractional flow reserve is very well reflected by the measured values in the same patients in the same regions. So it's interesting that this would not have been possible without introduction of artificial intelligence and finding the right algorithm to be able to, algorithm to be able to figure out exactly how to get meaningful data out of this study. Well, that's a very good point because we incorporate deep learning into our algorithm so that the more information we have with increasing knowledge from each subsequent CT scan, we validate that against measured FFR, and we can use that information to improve subsequent models. So it's a continuous cycle of improvement so that the accuracy of our analysis is always validated against measured FFR. And Stanford is well known for development of artificial intelligence in medicine. I think was one of the first institutions that actually initiated pretty extensive program in radiology, but in also in medicine in general. Is that correct? That's correct. And Stanford also has been a leader in imaging, in magnetic resonance imaging, and CT imaging, and introducing patient specific modeling. And that as I mentioned began with modeling of aneurysms and your men went on to the coronary arteries. Well, tell us a little bit about FFR, CT and interactive model results. So CT derived fractional flow reserve really begins with standard cardiac CT scans. These are the same CT scans that are performed every day in hospitals. The CT scan is uploaded through secure web-based interface to our servers of heart flow. Using special software which integrates artificial intelligence and deep learning, trained technicians developed a digital three-dimensional model of the coronary artery trees. We then apply the principles of computational fluid dynamics to compute coronary blood flow and to calculate fractional flow reserve at all points in the coronary tree. This is then delivered to the hospitals in several formats. One is a PDF format which can be printed. The second is an interactive format which can be accessed with your iPhone or iPad or on your hospital computer so that remote access and interrogation of these models are possible in physicians can now themselves see and evaluate the direct results of the analysis. The mean turnaround time is less than four hours to receive the results back in the hospital. So in this time of COVID-19, if you are not in the hospital working remotely, you can send a patient to the CT scanner and then receive the results of the analysis and communicate with the physician. This could all be then potentially remotely without in-person interaction with the patient. This is a simulation of what you may see on your iPhone. This is not an exact patient study, but it shows that you can click on the model at any point and you can derive an FFR value. The red color is a positive FFR less than .8 and the green blue and the green is a normal value. You can do pull back pressures on that and you can see significance of lesions. You can use those interactive models to plan treatments and to help guide decision-making. This example shows the anatomic functional mismatch that we sometimes see. There are two patients with 70% stenosis by CT. You can see the computational models. One of them shows a non-significant lesion with FFR CT of .86. This correlates very well with the angiogram and a measured FFR of .87. The second patient has a similar lesion on CT, but this is now a very positive lesion, color-coded red. You see the FFR value is .70. When that is compared now with the measured FFR value in the cath lab, it is .71. What looks like a pretty similar lesion on CT is very different on the functional level. Even if you look at these angiograms, it's hard to be sure just looking at the visual angiogram, which one is positive. When you measure FFR or compute FFR, the difference is very, very clear and it really can help guide patient treatment. This is very interesting because very frequently performing angiography, we misinterpret the severity of the disease. There is a tremendous variability between the operators on assessing the severity of the disease. I think this really exemplifies the importance of reliable and sensitive tests to give us the information which lesion is significant and which lesion is not significant. The other thing that's important is that in the computer at FFR, you get FFR values at all points in the corner tree. You can interrogate all the vessels from that one single model. When you measure FFR in the cath lab, you measure it at one point wherever the pressure wire happens to be in the artery. You can do pullback pressures, but it's such a more complicated, more time consuming procedure to interrogate the entire arterial tree with measured FFR. Very interesting. Tell us a little bit about diagnostic accuracy of a variety of non-invasive assessment of coronary disease. Of course, we have many different tools available. Now PET is very commonly used and it's thought to be very sensitive coronary, CTA, SPECT, and heart flow FFR. What kind of diagnostic accuracy do we have and how does heart flow FFR compare to the other modalities that we commonly use? That's a very important question. Diagnostic accuracy is really predicated on functional significance of lesions and that is as we've been talking about measured by FFR. The diagnostic accuracy of computed FFR compared directly to measured FFR was done in the NXT trial and that showed an 86% diagnostic accuracy. The other non-invasive tests such as PET, coronary CT, and SPECT have been tested against FFR in the Pacific trial run by Paul Knapin. And in that trial, in comparison to those non-invasive modalities with a sub-analysis comparing to heart flow FFR CT, we found that the highest diagnostic accuracy that is the highest area under the curve of an ROC curve was for FFR CT with an AUC of 0.94. This was significantly better than PET coronary CT or SPECT. We know that of all people going to the cath lab for diagnostic angiography, more than 50%, probably close to 60%, have normal coronary angiograms, that is non-obstructive disease. So in retrospect, you might say, well, why did they need to go to the cath lab if they don't have significant disease? So the same experience that we found in the platform trial now has been found in many other institutions in the world. That is the cancellation rate of angiography after FFR CT analysis. And overall, it's around 50% of patients. So that you can now reduce the negative catheterization rate in the cath lab by doing CT and FFR CT. So what does this mean? Well, this means that you can have more efficient use of the cath lab. So patients now are going to the cath lab not in order to get a diagnosis, but in order to get treatment. So if you identify patients ahead of time that they, in fact, have a significant lesion that needs to be treated, your average revascularization rate goes up. And that's, in fact, what happens if you look at these hospitals. And from multiple reports here, we see that without CT, FFR CT, the average rate of revascularization of all patients in the lab is about 34%. But once you start using CT, FFR CT, the revascularization rate is 69%. So that the patients in the cath lab are getting revascularized. This is particularly important in the era of COVID-19 because you don't want patients now in the high intensity cath labs. The patients themselves might contaminate the cath lab, or patients coming into the hospital might become contaminated because of all the COVID patients are in hospitals now. So that in order to improve the efficiency, safety, and effectiveness of the cath lab, pre-cath evaluation with not only an anatomic test like CT, but a functional test like FFR CT helps you sort out who doesn't, who does not need to be in the cath lab. So was this just in that one study, or is this a more generalized phenomenon in a real world experience, a registry, the advanced registry of more than 4,700 patients in 38 sites across the world, the same dynamics occurred. 63% of the patient management decisions were altered once FFR CT data was available. And importantly, and this is perhaps one of the most important observations here, is that patients who have FFR CT of greater than 0.8, that is they do not have hemodynamically significant lesions, even though they may have many, many plaques, they're not functionally significant. They do very, very well on medical treatment. Patients whose FFRs are less than 0.8, they have more disease, they need treatment, they need intervention, and overall, their cardiovascular death and MI rate is higher. But additional studies by Norgaard and Denmark show that those who now get treated with intervention do better. So this really helps decision making, helps guide treatment, and has improved outcome for patients. So what have we learned from all those studies that were carried on using CT and FFR? Well, we show that it's a really good diagnostic strategy. It's a good diagnostic strategy because you can exclude significant coronary disease, and many patients come in with symptoms thinking they have coronary disease, may have classic symptoms, but actually don't have significant disease, they have something else. The FFR CT analysis can identify which one of those patients with the disease has significant disease, ischemia producing coronary stenosis that may benefit from treatment. So you can differentiate patients who are best treated medically from patients who may benefit from a trip to the cath lab. The other important thing to learn here, especially in this COVID-19 era, is that this is a treatment strategy, a diagnostic and treatment strategy, that can be done with a single point of patient contact. That is a CT scan. You don't need second tests, you don't need hands-on interactions in high-intensity areas in the hospital, you can get a CT scan. This can be done even on an outpatient basis. And from the CT scan data, you can now get the functional information, you can sort out which patients have significant disease and which patients should be treated medically. And this can be done remotely if necessary. If the risk of bringing a patient into the office or into the hospital for consultation is too high because of the risk of COVID-19, you can communicate remotely. And if a patient does indeed have significant disease that is life-threatening, that patient then can be scheduled and brought into the high-intensity facility for treatment. So this is a benefit, reducing exposure to patients and health care personnel and provides more efficient utilization of resources. Personal protective equipment is lessened and we can use the cath lab for treatment rather than for diagnosis. All right, so let's talk a little bit about COVID-19 and the use of this test, how to streamline and triage patients who need intervention and patients who don't need intervention. And CMS also made some recommendations. Can you go through it a little bit with us, please? So the Centers for Medicare and Medicaid Services released guidance on March 17 asking health care providers to postpone adult elective surgeries and procedures including dental procedures until further notice. CMS recommended decisions about procedures to be made in consultation with a hospital surgeon, patient and other health care professionals. And the chart made up this, the agency made up this chart to guide decision-making. So you can see elective procedures are postponed, but you do not postpone high-acuity emergency procedures. These are acutely symptomatic patients. These are patients who may die if treatment is withheld. Importantly, perhaps patient with a left-main disease, patient with a critical ischemia-produced left-main disease which sudden death is not an unusual occurrence. So there are patients who on the basis of symptoms and presentations are urgent and should not be postponed. And of course, it's the physician's judgment to make the decision, does someone have such a lethal disease? There are exceptions for cancers. There are exceptions for patients with life-threatening conditions including significant coronary disease, which as you know is the number one cause of death in the world. So I think the CTA first pathway is a very important pathway because it allows definitive diagnosis of symptomatic patients who need cardiac testing. A CTA first strategy allows you to answer the first important question, does this patient have coronary disease or does this patient not have coronary disease? If the patient does not have coronary disease, that patient could be very well treated medically as appropriate. If the patient has anything between 30 and 90 percent stenosis, that could be a patient that you may want to get an FFR CT evaluation to determine whether or not this is hemodynamically significant ischemia producing. So this strategy can be done without excess contact with the patient other than the initial CT itself. So this can reduce the need for diagnostic angiograms, alleviate the burden to hospitals and health care facilities, and minimize negative cath procedures. It allows you to risk stratify patients, enabling you to select patients for revasterization and only referring patients who are at the highest risk to the immediate use of the cath lab with the remainder of patients can be deferred. Of course all of these decisions are made on the basis of doctor-patient interaction, communication, and the judgment of a physician based on the symptoms of the patient. So obviously I'm sure a lot of internationalists would have or those that take care of a patient with acute coronary syndrome questions for you related to several important issues that we have to deal with on a regular basis. Number one is how many institutions are able to perform good quality CTAs and how reliable those tests are to be able to get meaningful FFR data that will be obtained from the hard flow analysis. Well that's an important question because the hard flow analysis is dependent on a good high quality CT scan and the CT scan imagers of today are fully capable of producing high quality CT images. Now coronary, I'm not a cardiac CT expert but so you need to consult with them but a couple of things that are important. One is when you image for coronary CT for heart flow analysis is important to follow the strictest society guidelines including heart rate control. This will cause great improvements in image quality ideally heart rates of 60 or less are hoped for and to give coronary vasodilators such as nitrates just before the CT scan. The reason that is important is because coronary arteries that are not dilated will give an abnormal reading in the cath lab when they measure fractional floors or they give nitroglycerin vasodilators. So the comparison of FFR to computed FFR has to be on the basis of the same anatomy that is not constricted coronaries but dilated coronaries. So following SCCT guidelines for coronary acquisition the machines no matter which kind of 64 slice or greater machine you have is capable of good quality CT images. Now there are situations sometimes the patient moves. One feature that is talked about a lot is calcification. Now calcification interferes with image reading the quality of the images because of calcium blooming. It's been shown that FFR CT analysis is possible even in calcified arteries with good quality results. So care in CT imaging, training the technician, dedicated technician, following the protocols and careful attention to the process of image acquisition should allow most every place that offers cardiac CT to perform heart flow analysis if they wish. Now this could be a potential problem in the patient in the COVID era. A lot of those patients whether they have COVID or they don't have it but they have let's say cardiac symptoms and hemodynamic instability it is very difficult to control the heart rate and bring the heart rate down and they might have tachycardia or atrial fibrillation and rapid reticular response. They might be hypotensive where nitroglycerine cannot be administered so either we cannot do this study for that particular reason or it will be suboptimal. And is that of concern? Yes it certainly is and FFR CT analysis is indicated for stable patients not patients with recent MIs. If there's been a myocardial infarction within 30 days there's no evidence to show that this is useful. So this is for patients with stable symptomatic disease suspected coronary disease with stable symptoms not for acute patients so with acute instability being admitted with acute coronary syndromes. Well I understand that I was not trying to point this towards acute coronary syndrome but point towards the patient that admitted with possible COVID infection and symptoms. Symptoms could be due to many different reasons whether they are perbrile or have respiratory issues and so on. So I'm trying to figure out how we can make this beneficial and useful and safe in this type of a scenario. Well this is where the physicians need to get together to establish protocols within the hospital and how to manage patients and to prepare them properly for CT if they wish to perform CT. And this is a very important question that really needs to be looked at and addressed at each institution. One group of patients of course is in the emergency room. Patients come in with symptoms and may have symptoms that are very suggestive of acute coronary syndrome even though they are stable. And there need to be protocols in place. This requires the CT scanner being ready with trained technicians who can do proper quality CT scan and systems in place for data acquisition. Prompt reading of the study to determine whether or not there is a 30 to 90% lesion and then a system in place where those images then are uploaded to heart flow so that the results can be returned. So what else can we expect in the future as far as heart flow technology is concerned? Of course it takes up to four hours or maybe sometimes longer to get the reading and sometimes we don't have that luxury to be able to wait that long. So do you foresee in the future where artificial intelligence might lead us even further in being able to get this data in a shorter period of time? We are certainly working on that and the analysis times and the capabilities are getting shorter all the time. They will never get to zero of course but we are working on that in order to reduce turnaround time to get the reports back as quickly as possible. Certainly patients with relatively straightforward disease can be done more quickly than patients with complex multivessel calcified deletions. So that requires a little bit more time. But overall the process is getting faster. Well Dr. Zarans thank you very much for your very valuable contribution to Texas Heart Institute educational programs in COVID-19 epidemic. We are greatly appreciative of this information that you have provided to us. Thank you very much. Thank you very much and good luck to all the healthcare workers in this difficult time.