Hello, I'm Panitia Charantaitui from the Mayo Clinic. I'd like to thank the course directors and ASNIC for inviting me to participate in this cardiac PET workshop, and I look forward to seeing you at the live event. These are my disclosures. Over the next 30 minutes, I'm going to review metabolic imaging with PET. I'm going to provide a brief overview of FDG, then I'll discuss the use of FDG PET in the assessment of myocardial viability, cardiac sarcoidosis, and other types of inflammation, as well as cardiac infection. Most of you are very familiar with FDG, which is F18 labeled fluorodeoxyglucose. It's a glucose analog that's taken up by metabolically active tissue, which includes normal tissue where we may see physiologic myocardial uptake. We may also see FDG uptake in inflamed or infectious tissues, hibernating myocardium, cancer, and other conditions. How FDG works is that it traces the initial transmembranous exchange of glucose from blood into cells, and its subsequent hexokinase-mediated phosphorylation into glucose-6-phosphate. But unlike glucose-6-phosphate, phosphorylated FDG is neither a substrate for dephosphorylation nor for the glycolytic pathway, and it becomes trapped within the cell and allows us to image it. Depending on the patient's metabolic environment, for example, the levels of insulin and free fatty acids, the heart may take up more or less glucose and also more or less FDG. An important indication for FDG PET is the identification of myocardial viability in patients with ischemic cardiomyopathy, where we'll spend the most time. For this topic, I'll first provide a brief review of the background, then I'll discuss PET viability protocols, patient preparation, and interpretation of the PET images. We'll also examine the evidence for using FDG for viability assessment and provide some recommendations. So, first, some definitions. The two types of dysfunctional viable myocardium are hibernation and stunning, which in theory have distinct definitions. Hibernation is defined as a state of chronic and tractile dysfunction due to persistently low myocardial blood flow. It's postulated to be an adaptive response to maintain cellular viability. These adaptive mechanisms include a metabolic switch from fat to glucose metabolism and a reactivation of the fetal gene program. In hibernation, contractile function should improve by improving blood flow. Stunning, on the other hand, is defined as a state of contractile dysfunction due to transient ischemia followed by restoration of flow. So with stunning, function should just improve with time. In theory, these two myocardial states seem quite distinct, but in reality, they may be much less distinct and may coexist in the same heart or even in the same myocardial segment and may be part of a spectrum of the same process. In fact, studies using PET to quantify myocardial blood flow have shown that normal resting flow exists in hibernation, but it's really the abnormal flow reserve that's present, suggesting that hibernation is actually due to repetitive stunning. So both types of myocardial states represent dysfunctional viable myocardium that may potentially benefit from revascularization. So what are these benefits from revascularization? Based on observational studies, a subset of patients with moderate to severe ischemic cardiomyopathy who undergo revascularization may experience the benefits shown on this slide with improved survival, higher ejection fraction, less ischemia, less heart failure, among other benefits. But at the same time, these patients also have a higher risk of periprocedural and perioperative complications. So the rationale for viability testing has been to identify patients who may benefit from revascularization when they have moderate to severe ischemic cardiomyopathy. Traditionally, FDG PET has been considered the gold standard for the assessment of my cardial viability among the non-invasive viability techniques, because it has the highest sensitivity as shown here and negative predictive value for predicting regional contractile recovery after revascularization. How FDG works for detecting hibernation is shown here. In the resting state, the normal myocyte prefers to use free fatty acids for energy production. And so myocardial glucose uptake into the cell and proportionately FDG uptake are relatively low. On the other hand, the ischemic or hibernating myocyte prefers to use glucose rather than free fatty acids for energy production, leading to more glucose and also more FDG getting into the cell and becoming trapped for us to image. The traditional PET viability protocol may include only a resting perfusion image and an FDG PET image. But many centers now perform both a resting and stress myocardial perfusion imaging study before going ahead with the FDG portion of the study. Once completed, the images can be viewed to determine whether the patient actually needs to go on to the FDG portion of the study. For example, in this patient with severe multivessel disease and EF of 30 to 35%, these ammonia and 13 ammonia perfusion images show a large reversible apical, interoceptal, and lateral defect with only a minimal residual defect in three apical segments at rest. Here are the images again in color scale showing the large reversible defect with very small residual defect at rest. Because N13 ammonia requires an intact cell membrane for uptake, the uptake of N13 ammonia at rest is also a marker of viability. So with this, we conferred with the referring provider and together we decided not to go on to the FDG portion of the study given the presence of the large area of ischemia, the mild residual resting defect, and the patient discomfort during the two prior acquisitions. So instead of performing a four to five hour test, we were finished after an hour and a half. Contrast this with the second case of a young man with a recent cardiac arrest and EF of 30% and severe angiographic CAD who has a large severe fixed apical, inferior, interoceptal, and lateral defect without any stress-induced ischemia. He went on to have the FDG portion of the study to determine if he has additional viability in areas with perfusion defects since FDG has higher sensitivity than the SPECT or perfusion PET. To go on to the FDG portion of the viability study, we must perform metabolic manipulation to stimulate maximal FDG uptake in the myocardium to optimize image quality. Without the metabolic manipulation, FDG images would look like the suboptimal images in the middle row here where the FDG was injected under fasting conditions without any glucose or insulin stimulation. Contrast this with the bottom row where the same patient received glucose and insulin and the images are better optimized for interpretation and we can more confidently say that there is a perfusion metabolism mismatch in the apex and the mid to distal anterior wall. For myocardial viability assessment, the goal of the metabolic manipulation is to shift towards myocardial glucose metabolism to maximize glucose and FDG uptake in the myocardium. While approaches have been used as shown here on the slide, I would avoid the fasting protocol because that tends to give suboptimal image quality. Nicotinic acid derivatives are not approved in the United States, so the majority of practices use the oral glucose loading protocol, which is also recommended by the 2018 ASNIC practice point shown here. This protocol provides satisfactory image quality in the majority of non-diabetics and also does really quite well in diabetics if sufficient insulin is administered. The details of the protocol can be found on this table from the ASNIC practice points. On the ASNIC website, it's used in patients with a fasting glucose of less than 250. All patients should pass for 6 to 12 hours prior to the study. They're then given an oral glucose load and then insulin is administered according to the algorithm shown here. The goal is to maintain a blood glucose between 100 to 140 at the time of FDG injection. In patients with a fasting blood glucose of greater than 250 and or diabetes, no glucose load is needed and regular insulin can be administered just to bring down the blood glucose to 140. If upon imaging, poor target to background ratio is demonstrated, additional insulin may also be administered and the patient re-imaged after waiting another 45 to 60 minutes. As can be seen from this figure, the glucose manipulation can take 2 to 3 hours or more. And once the blood glucose is 140, the FDG is injected and there is an uptake period of another 60 minutes before image acquisition can begin. So there is really no easy way to speed this up. And the test typically takes us 4 to 5 hours. So it's really important to emphasize checking the perfusion images first to be sure that the FDG portion of the study is really needed. Because most of our patients are diabetics, we use the hyperinsulinemic euclasemic clamp protocol to ensure optimal image quality in all of our patients. And this provides a standardized approach, but it's time consuming and labor intensive. The goal of the clamp is to raise the plasma insulin concentration acutely to a new plateau of 100 microunits per mil as shown here by using a prime continuous infusion of insulin at a standardized dose for all patients at 1 milliunit per kilogram per minute. This level of insulin would generally result in profound hypoglycemia, but we maintain a plasma glucose at a euclasemic level that's arbitrarily defined here at 90 milligrams per deciliter shown here in purple. And we do maintain this by using a negative feedback principle using a 20% dextrose infusion. Viability image interpretation should incorporate all available sources of data, including the clinical history, other imaging data, echo, cardiac MR, and coronary anatomy if available to answer the specific question about viability in certain segments. And so interpretation of the PET data should include the combined assessment of perfusion, FDG, and regional contractile function. The normal myocardium should exhibit normal perfusion, homogeneous FDG uptake as shown here in the glucose loaded state, and normal regional function. There may be exceptions to this with left bundle branch block, for example, and ventricular pacing where the reverse mismatch pattern may be seen. And in that case, the septum should be viable based on the perfusion data. Hibernation exhibits reduced perfusion with relatively preserved or even enhanced FDG uptake. And this is the classic PET perfusion metabolism mismatch. And this is shown here in this example with the mismatch in the apex and also the mid-to-distal anterior wall. And these segments should also exhibit a wall motion abnormality. This pattern has the highest likelihood of contractile recovery after revascularization. Scar exhibits decreased perfusion, decreased FDG uptake, and decreased contractile function. So it's a match defect. And it has here is shown again in the apex and from the mid-to-distal anterior wall. When there is a severe match perfusion defect to this degree, the likelihood of functional recovery is quite low. Stunning may exhibit normal or slightly reduced perfusion and slightly reduced FDG uptake and has reduced contractile function. So there could be a perfusion contraction mismatch. And this pattern has somewhat uncertain likelihood of contractile recovery with revascularization. But there may be other clinical benefits to improving myocardial blood flow. So I've covered the majority of the strengths of PET for viability assessment. But here they are again on a single slide for you. The limitations of PET include limited access, cost, and expertise. And while PET has high sensitivity, remember that it has lower specificity for contractile recovery. Lastly, there's also radiation exposure to the patient up to about 15 millisieverts. Next, I'd like to show you a few slides on the evidence in support of using PET for viability assessment. I've already shown you this slide in which PET has the highest sensitivity for predicting regional contractile recovery after revascularization. And these data come from many observational studies involving thousands of patients. Again, note the lower specificity. In addition, based on observational studies, PET can also be prognostic and can help to guide management. In this single observational study, there are three key points. The first key point is that on the left, when viability by PET was present, survival was very poor with medical therapy as shown in green with a 50% annual mortality. The second key point is that again on the left, when viability by PET was present, revascularization in blue was associated with significantly improved survival. The third key point is that on the right, when viability by PET was not present, there was no significant difference in outcomes with medical therapy as compared to revascularization. So there are many more observational studies similar to this one that have shown other benefits of PET viability imaging in several thousand patients. But in the interest of time, I will summarize that there are limitations to the observational literature as listed on this slide. A few of them are the small sample sizes, varied protocols, variable duration of dysfunction before revascularization, and variable follow-up. On the other hand, there have been very few randomized trials on the role of viability testing in outcomes with revascularization versus medical therapy. One of the few randomized studies is the PAR-2 trial, which was conducted at nine North American sites, mostly in Canada. It enrolled 430 patients with severe ischemic cardiomyopathy. Patients were randomized to management assisted by FTG PET, as shown here, or standard care. Patients were then followed for cardiac events. The primary outcome was a composite of cardiac death, MI, or recurrent hospital stay for cardiac cause. At one year and at five years, the PET arm in red tended to have fewer composite endpoints and cardiac events compared to standard care in blue, but the difference was not statistically significant. Over a quarter of patients in the PET arm did not adhere to the trial recommendations based on the PET findings. When only patients in the PET arm who adhere to the trial recommendations were included in the analysis and compared to the standard arm, there was significantly greater survival in the PET adherence arm in red as compared to the standard arm in blue, both at one year and at five years. So it seems that if PET viability is performed and interpreted by experts and the results are used to guide management in a judicious manner, then outcomes may be improved than if PET were not used. The other study I'm going to mention is the stitch viability sub-study, but it's important to note that while the main stitch trial was a randomized controlled trial, the stitch viability sub-study was not a randomized part of the trial. And in the viability substudy, patients had optional viability testing by SPECT, dobutamine echo, or both. The viability substudy also did not include PET viability testing and had small numbers of patients in the non-viable arm. So unlike the prior observational studies, where we saw improved survival with revascularization, when viability was present, in the STICH viability substudy, there was no significant interaction between the presence or absence of myocardial viability and the beneficial effect of CABG over medical therapy. Although the trial was rigorously conducted, it did have some limitations, as I mentioned, by not being randomized for the viability substudy, it didn't include PET, and it had other limitations. So we likely need larger studies to confirm these findings more definitively. Based on all of the available evidence, it's likely that routine viability testing in all patients being considered for revascularization is not recommended, especially not in the younger patients, those with clear angina, which implies ischemia and viability, those with moderate to severe ischemia on other testing, and perhaps those with higher EF and left main and or no or limited comorbidities. On the other hand, viability testing may add useful information in older patients, in those without angina, without stress-induced ischemia, or if they have lower EF, chronic total occlusion or severe or multiple comorbidities. So now we're going to switch gears completely and talk about another very important indication for a metabolic FDG PET imaging, and that's cardiac sarcoidosis. I'll go over a bit of the background then discuss protocols and interpretation and review a few studies from the literature. So sarcoidosis has been in the limelight in the last few years, so this is likely a review for most of you. Sarcoidosis is a multi-system disease that can affect nearly any organ. Occupational, environmental, infectious, and genetic causes have been implicated. The central role in disease development is an exaggerated immune response, which manifests as intense inflammation akin to a brush fire, and over time, this leads to non-caseating, non-necrotic granulomas, as shown on this histopathology slide. Sarcoidosis is relatively rare. The disease has variable clinical presentation. Depending on the involved organs, the lungs and lymph nodes are the most involved in more than 90% of cases. The rate of cardiac involvement is uncertain depending on whether the report is an autopsy or an imaging case series. Cardiac involvement is often silent, further complicating the diagnosis. Mortality of sarcoid without heart involvement is variable, but generally low. It may be higher with cardiac involvement, especially in Japan, where the 10-year survival with reduced EF was reported to be only 19%. On the other hand, recent data suggests that 10-year survival with normal EF is nearly 100%, so there is now some uncertainty about the true prognosis of cardiac sarcoidosis. Cardiac sarcoid presents with a very broad range of clinical manifestations, from having no symptoms at all to having sudden cardiac death. Conduction system problems are the most frequent, followed by ventricular arrhythmias. The different manifestations are a direct result of where the granulomas and fibrosis like to go, and this can be in the myocardium, pericardium, or endocardium. But most commonly, sarcoid likes to go to the septum, especially the basal septum, and in this autopsy series of 41 cases of sudden death, probably due to cardiac sarcoidosis, the ventricular septum did show the greatest involvement, followed by other locations as shown on this graph. While it's possible to have diffuse cardiac sarcoid, it's quite rare. Advanced imaging is very important to the diagnosis and management of cardiac sarcoidosis because of the low yield of endomyocardial biopsy and also of echocardiography of only about 20 to 25%. MRI and PET both have unique strengths, but PET may have higher sensitivity and is the best technique for disease activity, i.e. inflammation, and is the only modality that can reliably provide information on both cardiac and extracardiac disease activity. The joint SNMI-ASNIC expert consensus document identified four clinical scenarios in which PET-CT may be useful in suspected or known cardiac sarcoidosis. These are one, patients with histologic evidence of extracardiac sarcoidosis and abnormal screening, defined as any of the items listed on the slide, two, patients less than 60 years old with unexplained new or new second or third degree heart block, three, patients with unexplained BT, and four, patients with proven cardiac sarcoidosis to help follow response to therapy. The PET protocol for cardiac sarcoidosis recommended by the SNMI and ASNIC is shown on this slide. A critical step to the success of this protocol is patient preparation to suppress physiologic myocardial uptake so that we can unmask FDG uptake due to inflammation and have high diagnostic accuracy. This involves consumption of high-fat, low-carb meals the day before the FDG PET, and in the US, this usually involves fried fatty meats and oil or butter, but fried eggs could also be used. Everything else must be avoided because everything else contains carbs. Patients must also finish eating by 5 p.m. if they're coming in the next morning so that there's a prolonged fast in conjunction with the fatty meals. The patients are also instructed not to exercise for 24 hours before the PET. We don't give adjunctive heparin, but this is given by many practices just before the FDG acquisition. Our nurse calls every patient in advance of the study to be sure they understand the importance of the prep and they answer any questions that patients might have. And using this approach, we have a 91 success rate of suppression that we've published for our cardiac sarcoid PET studies. The SNMI and ASNIC recommend that two sets of images be obtained to differentiate the spectrum of cardiac sarcoidosis. This includes the gated myocardial perfusion images and the cardiac FDG images. With the same FDG injection though, patients should also undergo a limited whole body PET study that should include the chest, liver and spleen where sarcoid likes to go. The limited whole body scan is also very useful for diagnosis, prognosis, possible biopsy sites and to help guide management. Image interpretation should be performed by simultaneously analyzing both resting perfusion and FDG images side by side. In the first row here, a normal study has both normal breast perfusion with complete suppression of FDG from the myocardium. This could also represent treated cardiac sarcoidosis. In the second row, there is no perfusion abnormality, but there is diffuse FDG uptake in the myocardium. This could be failure to suppress physiologic myocardial uptake of FDG or rarely, more rarely diffuse disease. The third and fourth rows demonstrate focal FDG activity consistent with active inflammation with or without an associated perfusion defect. The perfusion defects are set to be due to compression of the microvasculature from the granulomas. In the last three rows, there's FDG uptake with more significant perfusion defects. And here there may be some scar tissue involved. Remember that isolated FDG uptake along the basolateral wall without any perfusion defect is often a nonspecific finding. If there's a match defect of FDG, absent FDG uptake and also a severe perfusion abnormality that could represent scar. How is PET used then to make the diagnosis of cardiac sarcoidosis? Several societies have developed algorithms for the diagnosis of cardiac sarcoidosis. And the two most commonly used are the Japanese and the Heart Rhythm Societies. Both have a histologic pathway for the diagnosis of cardiac sarcoidosis which require non-caseating granulomas demonstrated on endomyocardial biopsy. But remember that EMB has very low sensitivity of only 20%. So most patients are diagnosed using the clinical pathway and each society has clinical pathways which are shown on the slide and are similar between the two societies. Both societies require a diagnosis of extracardiac sarcoidosis in addition to other criteria which include cardiomyopathy, high-grade AV block, unexplained BT, positive MRI or positive cardiac PET. And again, this is where FDG PET CT is particularly advantageous given the ability to readily identify not only cardiac but also extracardiac sarcoidosis. The Japanese society also includes diagnostic guidelines for isolated cardiac sarcoidosis shown here for your reference and is self-explanatory. How can PET help us with prognosis? Our colleagues at the Brigham reported the worst outcome in patients with both abnormal perfusion and abnormal FDG shown by the red line with a threefold increased risk in VT or death as compared to those without either abnormality. So here's another reason to perform the perfusion in conjunction with the FDG image for cardiac sarcoidosis. In the same study, focal RV FDG uptake was associated with a fivefold increased risk in VT or death as compared to those without focal RV FDG uptake. A limitation of this study is the small sample size but it's still very important to look for that focal RV FDG uptake when reviewing the cardiac PET images for cardiac sarcoidosis. Once immunosuppressive therapy has been initiated, repeat FDG PET imaging can be performed to help assess response to therapy in cardiac sarcoidosis. And for serial FDG PET imaging, it's very important to do everything in the same way. The SNMI and ASNIC also considered that change is more likely to be clinically significant when both intensity and volume of FDG uptake change in the same direction and by at least 20%. And for serial imaging, we can also look at SUV measurements before and after treatment to see if they may help us determine response to therapy. And we'll go over these a little bit more during the live sessions. Lastly, it's very important to keep in mind the nonspecific nature of FDG uptake and interpretation really requires meticulous and comprehensive review of the history and all the available data as well as patient preparation. Also consider the coronary anatomy, especially if there are significant perfusion abnormalities. And then lastly, the literature in this area is limited to small patient numbers, selection bias, as well as lack of standardization among different centers. Another indication for FDG PET CT is large vessel vasculitis. And there are several approaches to this and one of them is shown here. Grade two and three uptake where uptake is equal to or greater than liver uptake as shown here is considered to be possibly indicative and positive for active large vessel vasculitis respectively. And the sensitivity and specificity of FDG PET for large vessel vasculitis are really quite robust as shown on this slide. It may be helpful for vasculitis to go ahead and use the cardiac sarcoid prep protocol since there may be some coronary and myocardial involvement but the perfusion image is probably not necessary. Lastly, the use of FDG PET for cardiac infection has really emerged in the last few years because of the high morbidity and mortality of these patients, early diagnosis is very important. And to help increase the early diagnosis of prosthetic valve infection, where TEE as well as TTE may miss some of these cases, the ESC has included both FDG PET-CT and white blood cell SPECT-CT as major criterion in its diagnostic algorithm. The data supporting the use of FDG PET-CT and prosthetic valve endocarditis come from eight studies have been included in this meta-analysis. The studies are all relatively small with some limitations but they suggest that adding FDG PET-CT to the modified due criteria can substantially increase the sensitivity of detecting prosthetic valve endocarditis without losing much in the way of specificity. FDG PET-CT helps to overcome the limitations of acoustic shadowing by the prosthetic valve and allows earlier diagnosis than TEE and CT. Similar to what I showed you with cardiac sarcoidosis, it also provides an assessment of potential extracardiac infectious sources as shown on the slide where whole body FDG PET-CT showed a fourth toe osteomyelitis. It can also help identify embolic consequences of infective endocarditis and can help with patient management. For native valve endocarditis, the ESC didn't include PET-CT as a major criterion because based on a meta-analysis that included nine studies in patients with native valves, the sensitivity of FDG PET-CT for native valve endocarditis was very low. For pacemakers and ICT infections, FDG PET-CT does really quite well and shown are the overall sensitivity and specificity. But if we drill down to the different components, sensitivity is lower for device leads and higher for pocket infection. Studies have also shown that FDG PET-CT can help diagnose LVAD-related infections with very high sensitivity and modestly high specificity. But again, studies are limited. So here's a proposed diagnostic scheme by Drs. Wenken-Chen and Dilcisian for cardiac device infections. For evaluation of suspected valve prosthesis or lead-related endocarditis, TEE and CT are first-line imaging modalities. If TEE and CTA reveal a diagnosis of endocarditis, FDG PET-CT should be added for evaluation of extracardiac infection source or infectious emboli. If TEE or CTA are inconclusive, then PET-CT can also be performed to provide incremental information to confirm or exclude the diagnosis. For pocket and LVAD infections, FDG PET-CT should be first-line as their locations are generally beyond the ranges of TEE and CTA. If findings on PET-CT are uncertain or with a concern for a false positive diagnosis, a white blood cell scan can then be added. And here are the limitations of FDG PET-CT in cardiac infection. We still need to do the same complicated patient prep as in cardiac sarcoidosis, but we don't need the perfusion image. A major concern for FDG PET-CT for infection is the nonspecific uptake of FDG in sterile inflammation triggered by synthetic grafts or implanted devices. So FDG PET-CT is generally not recommended within the first three months after surgical placement. Then for cardiac infection, SUV measurement hasn't been shown to be able to differentiate infection from inflammation. So their use is generally not recommended in cardiac infection. Then lastly, the literature has similar limitations to cardiac sarcoidosis literature without any randomized controlled trials and with small sample size and other limitations. This concludes my presentation. Thank you so much again for including me and for your attention. Thank you.

metabolic imaging

PET

FDG

myocardial viability

cardiac sarcoidosis

inflammation

infection