Hello, this is Brian Young from Yale University School of Medicine, Division of Cardiology. This is Module 5B, Pet Tracers for Myocardial Inflammation, Viability, and Amyloid Imaging. This module was developed in collaboration with Dr. Allison Neusses, also at Yale University. These are our disclosures. I have no disclosures pertinent to these topics. Our learning objectives today will be to focus on the principle and protocols of FDG-PET in three contexts, first in sarcoidosis imaging, in myocardial viability imaging, and lastly in the imaging of cardiac infections, both intracardiac and cardiac device infections. And the fourth item of our agenda today will be becoming familiar with available tracers in cardiac amyloidosis that are coming online clinically. This is an overview. PET tracer applications in the cardiovascular space can be broadly divided into three major groups. By far, the most broadly utilized of the PET tracers in cardiology are those used for perfusion imaging, which will be covered in separate modules. The second group of applications is in metabolic imaging, imaging in which FDG is used in a variety of protocols for different purposes. And the third group would be essentially everything else. And today that category will be represented by amyloid tracers that are presently coming online for cardiac amyloid imaging, many of which were borrowed from the central nervous system imaging space. So let's start with cardiac sarcoidosis today. Of note, a lot of the clinical subtleties and aspects of the application of these tracers and imaging results will be discussed in a longer separate module dedicated to sarcoidosis. But I would like to highlight when we employ this particular strategy, because it does make a difference to understanding what the value of the imaging is. So FDG cardiac PET should be considered in the context of other multimodality imaging when there's an abnormal screening for cardiac sarcoidosis. And by that I mean in patients who have histologic evidence of extracardiac sarcoidosis, most typically pulmonary sarcoidosis, when they come to attention based on an abnormal cardiac finding during routine clinical monitoring, such as abnormalities on ECG, on Holter monitoring, or other clinical criteria. So many of these patients will have established systemic sarcoidosis and will come to clinical attention because there's increased ectopy on a Holter monitor or vent monitor that's done for a complaint of palpitations or something else, or that there's evidence of new conduction disease on a routine screening ECG clinically, or if there are other clinical criteria that have come to attention because of other cardiac imaging that has been done, such as MRI or echocardiography. The other groups outside of the group of cardiac suspicion patients who have come to attention because of extrapulmonary disease are patients younger than 60 with unexplained new onset conduction disease who are being investigated for the etiology thereof, or in patients with idiopathic sustained ventricular tachycardias. And finally, one of the major groups that we use this kind of imaging in is patients with proven cardiac sarcoidosis, these are known cardiac patients who are being followed for the progression of their disease or response to disease therapy, such as immune modulating agents. Of note, the Heart Rhythm Society has added FDG positivity to its diagnostic criteria for cardiac sarcoidosis, and also of note, endomyocardial biopsy, which you might suspect as being the gold standard for an infiltrative cardiac disease is actually of low sensitivity because of the patchy nature of cardiac sarcoidosis. Thus, in practice, FDG PET and cardiac MRI have become very essential tools for both the diagnosis and the monitoring of disease progression and therapy response. Now, to understand why FDG can be used in cardiac sarcoidosis and understand the specific protocols needed for it to be diagnostic, you have to understand the underlying metabolism that is used in this modality. The underlying principles of the cardiac myocyte has the capacity to use cellular energy both from glucose, from plasma taken up by glute transporters, or to drive the Krebs cycle by free fatty acid uptake and oxidation. So the goal in the prep for cardiac sarcoidosis is not to starve the cell of glucose because this isn't possible, it's a tightly regulated substrate, but rather to create a relative decrease in insulin and an increase in serum free fatty acids that will result in a down regulation of cell surface expression of myocardial glute transporters, favoring an intracellular shift toward the free fatty acid pathway. Lower glute transporter activity in the cardiac myocyte is what creates a contrast between the myocardium, which will then have a limited capacity to uptake the FDG tracer. Meanwhile, the uptake of FDG will be preserved in the cells of the immune system where glute three is a constitutively expressed transporter for glucose and is not affected by the free fatty acid insulin conditions of the systemic circulation. Thus you can use this contrast that you've generated through dietary prep to identify sarcoid granulomas. Also as we'll see later, leukocyte glucose metabolism is used similarly to image infection associated inflammation, as opposed to granuloma. So this is a schematic representation of the strategy of manipulating and then imaging the differences in myocardial metabolism in contrast to inflammatory cell metabolism to identify granulomas in CS. In contrast to perfusion imaging, FDG PET for cardiac sarcoidosis is a hotspot imaging technique. In the normal heart, fed a normal diet inclusive of carbohydrate, the myocardium takes up FDG and the image looks similar to a normal myocardial perfusion image. However, with a no carbohydrate diet, which I'll outline shortly, insulin is suppressed and circulating free fatty acids are markedly increased, resulting in a shift toward lower myocyte glucose uptake and clearing the way for inflammatory cells specific FDG uptake is quantifiable and of diagnostic utility. So again, the central concept here is that this metabolic manipulation going from this to this permits the imaging of this patchy FDG uptake that indicates the presence of an infiltration of granuloma into the myocyte, into the myocardium. As you might imagine, looking at the strategy, the dietary prep to achieve these metabolic shifts is really critical to getting the imaging to work. There are some differences in the overall approach to this protocol, but the backbone of the protocol of prep is the same. I'll emphasize here that in the experience of our center, the specificity of the instructions and the consistent direct patient contact, such as by phone calls by nuclear lab staff ahead of the tests is really the key to improving the diagnostic quality of these tests and avoiding a low quality or even a non-diagnostic scan secondary to high myocardial background. So here's how it works. The patients fast after their evening meal, and this evening meal is the conclusion of a day of a high fat near zero carbohydrate diet as best as can be achieved. And what I'd emphasize here is that specific lists of do's and don'ts, as opposed to giving a patient general macronutrient categories is really the key to success. Give the patient a menu, fried eggs, bacon, meats, fish, specifically outlining things that will sabotage the tests that are high in carbohydrates, even things that are high in carbohydrates that may otherwise be hidden from the patient's immediate consciousness here. So this diet will include both food and beverage, specifying very specifically that beverages are to be without sugar, ideally water, meds can be taken with some notable exceptions that I'll highlight in a moment. And the fast is critical overnight, leading into a test in our center is typically done scheduled first thing in the morning for the sarcoidosis patients. On the morning of the exam, they eat no food, they drink typically water and absolutely no carbohydrates again. So when it works and it works well, it looks like this. So this is a case of a 25 year old female who fasted for more than 18 hours after a heart high fat, low carbohydrate diet prep, similar to what I outlined. While glucose was normal, free fatty acids are markedly elevated. Myocardial uptake, as you see in these panels is notably suppressed. Very different than it would look say on a normal diet type FTG scan, such as might be done for cancer imaging of the heart would look much like the brain, very intense homogenous signal uptake. However, in this case, the heart is appropriately suppressed by the diet protocol. And what you see is focal patchy uptake within the heart, very consistent with granuloma with consistent with sarcoidosis. I will also point out that this case highlights why whole body PET imaging is best practice not only for oncologic imaging, but for sarcoidosis imaging, because sarcoidosis is a systemic disease and you see uptake in inguinal nodes, in the abdomen, in the thorax, out in the lung parenchyma, the mediastinum, in addition to what we see in the heart. All of this might have clinical utility. And so you don't want to sacrifice imaging that you could otherwise obtain with the same tracer dose you've already given. When it doesn't work well, it looks like this. This is a case of a 73 year old male with an inadequate dietary preparation. Note the uniform rather than focal or patchy uptake in the entire left ventricle. This indicates that the myocardium has not transitioned to free fatty acid metabolism. If reported in these two cases, you would actually see that insulin levels were markedly suppressed in the first case and not suppressed in the second. And the blood glucose, while normal in both cases, is in contrast to the fact that in the first case, the free fatty acids are markedly elevated in the second case, they're not as elevated. So this is a non-diagnostic study. If there were granulomas within this myocardium, you just wouldn't know because of the high background of the myocardium itself. So this is a test that would need to be repeated with better dietary prep or substituted with an alternative imaging modality such as cardiac MRI. The non-invasive imaging approach to initial evaluation of patients with suspected cardiac sarcoidosis is represented by this scheme here. And the purpose of this is to identify coexistent inflammation. So many centers will, as we do, approach this as a CMR preferred first test, as CMR is most closely associated with outcomes in terms of ventricular tachyarrhythmias. And if normal, meaning that there's no delayed gadolinium uptake to indicate the presence of myocardial inflammation or scar, these patients can be followed clinically. FDG PET serves as an adjunct to the MRI scan either when MRI cannot be pursued, when MRI is non-diagnostic or equivocal, or in cases where the diagnosis of sarcoidosis is made by MRI, but subsequent studies are focused on monitoring for active inflammation or the lack thereof in the case of a response to immunosuppressive therapy. This is an example of the utility of PET response in response to immunosuppressive cardiac sarcoidosis. A whole body FDG was done, I'm showing you here just the thoracic windows, and a patient without a response, you see persistence of the myocardial FDG uptake indicating ongoing presence of inflammatory cells. In a patient with a very marked response, you see that in both cases, the myocardial prep was sufficient in that there's not diffuse myocardial uptake, but just very extensive patchy myocardial uptake in the short and long axis views. But following treatment, there's complete loss of this myocardial signal indicating a response to therapy. This has prognostic power, I'll emphasize that FDG PET for this indication is performed with rest perfusion imaging, not something I showed in the previous slides, but always done most commonly in conjunction with rubidium 82, the PET perfusion tracer. This allows a separate assessment of myocardial scar as may occur late with myocardial granulomas and chronic disease, and together the presence or absence of scar and the presence or absence of FDG uptake have additive prognostic power as shown here in this study published in 2019 in Jane C. Normal perfusion FDG carries a good prognosis, abnormal perfusion or FDG is worse and then the worst prognosis is represented by this group, is carried by this group that has both abnormal perfusion and abnormal FDG indicating both the presence of ongoing inflammation and of scar, scarring of the myocardium. Lastly, I'll just briefly point out that uptake in the right ventricle is a particularly bad prognostic sign and PET has the capacity to identify these granulomas in the right ventricle with good spatial resolution and sensitivity in that area. This can be added, this could have additive value to the clinician who's trying to decide on potentially risky immunomodulatory therapies. So let's move on to FDG PET and myocardial viability imaging. The precision in the terminology is really essential to understanding the biology and the clinical application of this imaging modality. Viable myocardium includes normal myocardium, ischemic myocardium and hibernating myocardium. And the concept of hibernating myocardium was reviewed very nicely just a few years back in the reference shown here. And hibernation is really a metabolic adaptation to reduced flow or ischemia that is thought to maintain cellular viability in the setting of low flow states. I'll point out here that hibernation, the idea behind imaging it is that if identified these territories of hibernating myocardium are thought to be able to be brought back online and to be functional myocardium and contribute to myocardial mechanics and reduce poor prognosis if they're revascularized, thus the impetus to identify when hibernating myocardium is present. This preparation for hibernation is in marked contrast to the prep that we discussed for cardiac sarcoidosis. In this scan, a six to 12 hour fast is followed by a glucose load to induce a hyperinsulinemic state and standardize the cardiac substrate and increase FDG uptake in the myocytes, which is what you're trying to image in this case, in order to improve image quality. So what you're looking for is an FDG perfusion radio tracer mismatch. That is a metabolic uptake, an uptake of FDG that is preserved at the location of a perfusion defect. The perfusion is typically done either within 13 ammonia or rubidium 82. This has good spatial resolution and good attenuation properties based on the fact that these are all high energy photon PET tracers. So the best way to understand this is really to look at a clinical example, see how it works in practice. This is a case demonstrating both mismatched and matched perfusion defects within a single patient. On the initial stress and rest rubidium study, there's no ischemia. All of the defects that you see, which are quite severe and noticeable in this normal perfusion study here is, these are scars. There's no reversibility between stress and rest in a normal Ragged Denison PET protocol. On the right, rest rubidium images, instead of having rest and stress images, resting rubidium rubidium images are compared to FDG images. And what you see is that there's a match defect in the LAD territory in the anteroseptin. And to some extent, the anterior wall where an area of scar lacks FDG. So this is seen with the green arrows. Notice how this rubidium PET matches the FDG PET. There's a sizable mismatch, however, in the RCA territory, here pointed out by the orange arrows. There's a robust FTG uptake in the territory, in the inferior wall, where there's a marked perfusion defect as seen on the resting rubidium image. This is consistent with hibernating myocardium in the RCA territory and a potential clinical benefit to revascularization in this territory, while there is no apparent, by this imaging, viable or hibernating myocardium in this deep scar in the LAD territory. Patients with viable myocardium, as depicted in the prior slide, have decreased mortality with revascularization, so if viable, the death rate is improved by revascularization relative to medical therapy alone. However, in this 2012 meta-analysis, no difference was seen between, no benefit was seen in revascularization in the group that had non-viable myocardium. There are a lot of subtleties to the interpretation and clinical application of this data that are beyond the scope of this talk. However, I'll point out that FTG PET, in particular, among the viability imaging strategies, has been illustrated to identify people who have decreased mortality with revascularization, as seen in the previous graph, and with a composite of cardiac death, myocardial infarction, recurrent hospital stay for myocardial cause, as published a few years back in Jack, patients with revascularization did better, particularly if guided by a viability imaging guided strategy to applying revascularization. For other patients in this PARS-2 study, the differences in outcomes between FTG PET guided therapy and standard therapy were not statistically significant. However, another group undertook an Ottawa V sub-study, in which the proportional hazards regression showed a benefit for the PET assisted strategy, with a hazard ratio of 0.34. Thus, the clinical utility of viability testing may depend in part on the experience level of the cardiac PET laboratory, where the patients are being imaged, as well as the degree to which the downstream clinicians followed the recommendations of the viability imaging. In addition to the clinical data describing the utility of viability imaging, in terms of outcomes, FTG PET uptake, FTG uptake fits with the observations from other imaging modalities. To this point, what I've shown here is a correlation between the area at risk as defined by the defect in perfusion, here using Sustenibi, and FTG uptake, with the correlation statistically quantified below. I've not shown all the plots here, but the FTG extent also correlates well with LGE extent, as shown in the images, but not shown in the correlation plot. FTG extent appears and performs similarly with respect to change in LVEF, and as shown in the next slide, EF recovery. I'll also point out that this brings up an important caveat to FTG imaging in this setting, that is that FTG uptake is increased in the immediate post-infarct period because of inflammation, so that's actually what's being imaged here. This is a recent infarct in FTG going to the recent infarct because inflammatory cells are going to the recent infarct. This will become important in the subsequent section regarding FTG imaging and infection. Finally, here's shown an overlay of an anterior wall MI in the top panels and an inferior wall MI in the six bottom panels. More FTG uptake means less recovery of LV function and more dilation of the LV, and this is similar in predictive value as late gadolinium enhancement. The intensity of the FTG signal was also associated with MRI measures of LV global and regional functional outcome independent of the infarct size or the peripheral markers such as peripheral white blood cell count. And finally, FTG-PET has the highest sensitivity and negative predictive value of the available viability imaging modalities with specificity similar to that of these TEC-99 and thallium strategies, exceeding that of thallium, actually. So I'll shift gears a bit and talk about the only topic that doesn't involve FTG of these PET tracers in this module, and that is the tracers for cardiac amyloidosis. These are not yet widely used in clinical practice, but they are used in clinical practice for cardiac amyloidosis. These are not yet widely used in clinical practice to the extent of FTG and perfusion tracers, and they're not as widely used even in amyloidosis as the spec tracers PYP or DPD. However, there's abundant ongoing research and interest in developing these agents to improve multi-organ detection of amyloidosis, which is not only a cardiac, but a systemic disease, and to improve the ability to track and quantify progression and response to therapy of this disease, and to detect a difference in amyloidosis subtypes. In much of this, detecting subtype differences, assessing responses, is really suboptimal with the currently widely deployed spec tracers and existing imaging strategies, thus the motivation to push these tracers into the clinical domain. So, PYP and DPD are widely available and show excellent test characteristics with respect to diagnosis for TTR, cardiac amyloidosis. However, a negative test with those tracers does not exclude AL, and absolute quantitation is quite difficult with those tracers, limiting the ability to assess response to therapy, as I mentioned. So, in the last few years, there are new, exciting therapies available for TTR amyloids, including cardiac amyloidosis. Thus, there's interest in applying these PET tracers, many of which were developed for detecting amyloid in the central nervous system to cardiac imaging. These compounds have some structural similarities to thioflavin-T, a fluorescent dye that has been used since the 1980s to detect and actually quantify amyloid fibril formation, particularly in the research space. F18-labeled sodium fluoride is not shown here. Like PYP and DPD, it's also a calcium-binding tracer that differentiates TTR and amyloid patients, but it does not image AL, because TTR amyloid imaging is particularly sensitive to calcium imaging tracers. Thus, like SPECT tracers, F18-labeled sodium fluoride is of not of any differential particular clinical utility over the tracers that I'll be highlighting today, such as in differentiating cardiac subtypes or therapeutic response. So, I'll go through these rather quickly, as there are marked similarities between them. This is excellent research from Dr. DeBala's group, showing for beta-pyr binding specifically to myocardial light chain and trans-therapy enamel deposits, here shown in autoradiography and in histology. For beta-pyr is unique among these tracers in that it binds AL with higher affinity, as shown here in the lower right, AL, TTR, and control, and thus may identify AL earlier in marked contrast with our routinely used SPECT tracers, as well as assess other organ involvement. Here's an example of some intense myocardial uptake. On the left panels, you'll see that for beta-pyr is also useful in identifying pulmonary involvement. And on the right, you see for beta-pyr uptake specifically in the heart. There's data that for beta-pyr is unfortunately not able to reliably assess response to therapy, however. Next, we have a flutametamol. There's conflicting data for beta-pyr uptake, there's conflicting data regarding the diagnostic accuracy of this tracer for cardiac amyloidosis. This may relate to differential uptake patterns depending upon the amyloid pattern and fiber pattern. So, while these specificities are certainly of research interest and may ultimately have find their way to clinical utility, flutametamol is not in use at this time for cardiac amyloidosis outside of the research space and remains a tracer that you will see some data on in the central nervous system. Finally, for beta-pyr, this is illustrating T1-weighted late gadolinium-enhanced MR, and then whole body scintigraphy here with DPD, a bone tracer, calcium tracer, and showing how the florbetapin can image specifically cardiac uptake and how it correlates with the other tracers in some cases, but then in other cases such as in AL, we see that the AL, as I mentioned, is not identified well by the calcium-binding spec tracers such as DPD and PYP. However, it is successfully identified by florbetapin. So, in the lower right, you see that intense tracer retention in the myocardium. Florbetapin is thought to uniquely offer a single modality to not only diagnose cardiac amyloidosis but to differentiate between TTR and AL based on the degree of uptake. Genovese et al. actually showed in jack cardiovascular imaging that AL has a high persistent cardiac uptake using florbetapin, while over a period of 60 minutes, the early positive signal TTR washes out. Thus, its utility in differentiating not only the diagnosis of these diseases from normal patients, but the differentiation of the two diseases, TTR and AL, which have very different therapies. Finally, we'll move back to FDG, this time for identifying cardiac inflammation as evidence of cardiac infection, both in terms of intracardiac infection and cardiac device infections. In addition to the more commonly recognized use in prosthetic valve endocarditis, FDG PET has utility in PACER and ICD infection, both in terms of lead infection and in device pocket infections, as well as infections of other intracardiac devices, such as LVADs. Here's an example that demonstrates the excellent sensitivity of FDG PET and the detection of intracardiac infection. There are foci seen here of abnormal FDG uptake around the aortic prosthetic valve, suggesting an infection. In this case, interestingly, there were no abnormal findings on CT that correlate with this infection identified or this infection CT that correlate with this infection identified or this inflammation identified by PET, and even open surgical exploration. However, eight days later, after persistent bacteremia, it was noted that CTA showed persistent mycotic aneurysm, as shown at the arrow, and also an abscess on a transesophageal echo was seen. Repeat surgery showed severe infection. Again, highlighting that when other imaging modalities did not yet show more anatomic changes consistent with the developing infection around the valve prosthesis, FDG showed early evidence of inflammation in that area. This is FDG images from a patient with a native bicuspid aortic valve and a 28 millimeter gel weave ascending aortic arch graft, demonstrating increased FDG uptake in the ascending aorta down to the aortic valve in axial views, coronal and sagittal views. In this study, looking at PET and the diagnosis of prosthetic valve endocarditis, increased valvular FDG uptake was evaluated as a novel major criteria in the context of the Duke criteria in 72 patients. In the upper right is the performance of the FDG PET by final diagnosis, notably with 73 percent of patients that had a positive PET ultimately having definitive endocarditis, and 80 percent of patients without FDG signal ultimately rejected for endocarditis. And below is shown the performance of the Duke criteria with and without FDG as an added major criterion. In general terms, there's a reclassification from possible, a large group when using the Duke clinical criterion alone, into a possible or definite, the reclassification rather from possible to definite or excluded, increasing the sensitivity in particular at the expense of a modest decrease in specificity. This is a similar study including both prosthetic valves and intracardiac devices showing a reclassification of 90 percent, notice 50 patients in the classification of being possible rather than definitive or ruled out, 50 percent, rather 50 of these patients, 50 patients became five. In other words, a large reduction in the patients classified as possible infectious endocarditis by clinical criteria alone. For prosthetic valves, the test characteristics for both SUV max and for prosthetic material to background SUV ratio showed good sensitivity at the cutoffs shown, but also notably 100 percent specificities at higher cutoff values. In contrast, cutoff values were not useful in patients with devices because they were too close to the patient's background metabolic SUV activity. The central idea here is FDG is nonspecific in that it adds sensitivity but does not improve specificity and tracers in the future may better differentiate infection from inflammation. This differentiation is desirable because it is the, it's the nature of recently implanted prosthetics, be it in valves or be it a intracardiac device such as an ICD or pacemaker to have sterile inflammation around the device, which sterile inflammation is as imaged by FDG as would be an infection. So future tracers may actually afford us the possibility of doing imaging without the patient burden of the FDG prep, the dietary prep which is similar to that of the cardiac sarcoidosis prep, as well as eliminating that failure rate that is accompanied by not being able to achieve the prep. So here are two tracers that, one tracer rather, as compared to FDG that may actually improve the specificity for infection over non-infectious inflammation, and that is an F18 maltohexase PET tracer that is still in research development. Finally, in summary, FDG PET is used for the diagnosis and importantly for the monitoring of active disease and cardiac sarcoidosis. This requires a dietary prep that decreases myocardial FDG uptake in order to facilitate inflammatory cell imaging. The same strategy is used for inflammation imaging to identify infection, that is blanking the heart FDG uptake using the same dietary prep so that inflammation such as that around devices, around prosthetic valves, and around LVADs can be identified in contrast to a myocardium. And then viability imaging is a different use for FDG and an entirely different prep in which myocardium that is utilizing FDG specifically, being hibernating myocardium, is identified for the purposes of guiding reperfusion strategies. And lastly, we reviewed a few of the amyloidosis tracers that are in varying degrees of clinical deployment and development that have differential affinities for TTR and amyloidosis, which are diseases that are treated very differently. ASNIC has a variety of guideline documents and resources available to clinicians who are either looking to employ these tracers in terms of referral of their own patients or for cardiac imaging physicians who are looking to set up or augment their own imaging practice.

PET tracers

myocardial inflammation

viability imaging

amyloid imaging

sarcoidosis imaging

infection imaging

cardiac amyloidosis imaging