Hello everyone. Welcome to the PET curriculum series. My name is Krishna Patel and in my talk today I am going to provide you with a clinical overview of cardiac PET myocardial perfusion imaging, including myocardial blood flow quantitation. I have no relevant disclosures. And these are my learning objectives. Let's first start with what PET imaging is based on. In PET or positron emission tomography, we utilize positron emitting radio tracers. A positron is a positively charged electron which travels a short distance before it collides with an electron. This distance is known as a positron range and the collision event is called an annihilation reaction. This is when all the mass of a positron gets converted into energy and two high energy 511 keV gamma rays are emitted in opposite directions, which are then detected by a circular ring of detectors surrounding the patient present in a PET scanner. And then later these coincidence photon pairs are used to process and form an image. In contrast to a step and shoot acquisition with SPEC where the camera head rotates, a PET scanner has a circular ring of detectors which are stationary and do not move. And this is needed for simultaneous detection of coincidence photon events in opposite direction. Currently there are four PET radio tracers which can be used for PET perfusion imaging. Rubidium-82 is the most commonly used tracer in the United States. Its production requires purchase of a generator. Typically these generators have a four to six week cycle and can get expensive if there's not enough patient volume at a center to support its use. N13 ammonia is the next most commonly used radio tracer. Its use requires an on-site cyclotron for production. O15 water is not currently FDA approved and it's typically used for research purposes. And F18 fluorpyridase is a newer radio tracer which is currently in phase three clinical trials and not FDA approved. However it does have the potential to change the field significantly as it can be ordered as unit doses due to a relatively longer half-life. As you can see here that these radio tracers have very short half-lives which necessitates the use of pharmacologic stress test with it. Exercise is possible with N13 ammonia as well as fluorpyridase. However the downside of that is that myocardial blood flow quantitation is not possible with exercise stress. Not only that the shorter half-lives of these radio tracers also result in a much lower radiation dose to the patient with these SPECT radio tracers compared to SPECT. The other advantage of these SPECT perfusion radio tracers are they have very high, they have high first pass myocardial extraction fraction compared to technetium labeled radio tracers. Now what does that mean? Let's look at this figure on the bottom right here. Ideally we want our myocardial tracer uptake to be directly proportional in this 45 degree line with increase in coronary blood flow to the myocardium. So the higher the coronary blood flow the higher the tracer uptake. However you can see here that the technetium based radio tracers that are used with SPECT, the tracer uptakes plateaus relatively early on with increase in blood flow to the myocardium. Compared to that the PET perfusion radio tracers that you see here highlighted in red have a higher myocardial extraction fraction. Now what does that mean? It means that there will be higher counts within the myocardium and it makes it more easy to detect perfusion defects that are milder in severity and provides better defect resolution. Not only that it also allows for more precise blood flow quantitation as well. All of these characteristics contribute to a much superior diagnostic performance of PET compared to SPECT. What are some other key differences between SPECT and PET MPI? The spatial resolution with PET is around four to six millimeters up to two times higher than that of SPECT. All PET is attenuation corrected either with CT transmission scan which is performed in PET CT scanners or using a radionuclide source in a dedicated PET scanner. The timing of stress imaging in PET due to the short half-lives of these radio tracers is at peak stress unlike SPECT which is typically 30 to 60 minutes later. PET also allows for routine myocardial blood flow quantitation in all studies. Now all of these properties of PET that we talked about, the higher resolution, the attenuation and scatter correction and higher count density contribute to a much improved image quality with PET compared to SPECT. We can see here in this example of a 49-year-old obese male with multiple cardiovascular risk factors that the non-attenuation corrected SPECT shows an apparent inferior wall reversible perfusion defect which is possibly from the increased sub-diaphragmatic activity that you see here. While on attenuation corrected SPECT the inferior wall is corrected and you see on the rest image due to over correction there is an apparent anterior wall defect. On the same patient in PET you see that the images are great and you can say with confidence that this is a completely normal study. Within PET the image quality depends on the positron range of a radio tracer as well as its first pass myocardial extraction fraction. The shorter the distance that the positron has to travel before it annihilates the better the resolution. So both ammonia as well as fluoropyridase actually have shorter positron ranges compared to rubidium-82 which leads to better resolution with these radio tracers. They also have a higher first pass myocardial extraction fraction which leads to increased sensitivity and defect resolution like we talked about. It's important to note that O15 water the relative perfusion images are uninterpretable due to poor count density and typically we use that for myocardial flow quantitation. Now we've talked about how PET radio tracers result in a much lower radiation dose to the patient compared to SPECT but let's look at how much. The average radiation associated with a typical rubidium stress study is about four millisieverts that is that with a ammonia study is about three millisieverts and a fluoropyridase study is about six millisieverts. All of this is significantly lower than the estimated effective radiation dose from a one-day rest stress technetium SPECT which is about 11 millisieverts. In this study where they analyze the effective radiation dose from all studies that were submitted to IAC for accreditation the median effective estimated radiation dose for PET was 3.7 millisieverts from about 532 patients from 111 labs much lower compared to the median radiation dose from for SPECT which was about 12.8 millisieverts per study from about 3000 patients. Because of this ASNIC recommends to consider PET when it is available and that the study is appropriate and indicated as one of the key measures to reduce radiation exposure associated with myocardial perfusion imaging. Now let's look at the key indications for PET myocardial perfusion imaging. They are described in detail in this publication referenced here and I encourage you all to read it. PET is the preferred test in all patients who are unable to exercise or complete a treadmill stress test and those who require a pharmacologic stress test. However it is recommended in all patients who've had prior poor quality stress imaging. These are patients who've had an equivocal or inconclusive stress test result in the past. Those who've had significant attenuation artifacts on their studies and those who have discordant results on between the MPI study and and other imaging tests or CAF. It is also recommended for patients who have body characteristics that can affect their image quality. Now these are patients who who are either obese, those who have large breasts, chest wall deformities, pleural effusions, basically anything where you would expect the specced images to be suboptimal. It is also recommended in high-risk patients where you want to try and avoid diagnostic errors as much as possible. Now these are patients such as those who have chronic kidney disease or diabetes. Also in patients who have suspected left main or multi-vessel disease, those with extensive coronary artery disease who are symptomatic where you want to know if there's a new ischemia or not. Transplant patients where you suspect transplant vasculopathy and patients who are undergoing high-risk revascularization procedures to try and guide their revascularization decisions. That is also recommended in young patients who have established coronary artery disease. These patients typically require multiple procedures with radiation throughout their lifetime so you want to try and minimize the radiation exposure associated with non-invasive testing. And finally PET is also recommended in patients where myocardial blood flow quantitation is very useful for clinical decision making. These are patients where you suspect microvascular dysfunction or those patients where you suspect extensive multi-vessel disease. This is a typical rest stress PET rubidium PET protocol. We first start with a CT topogram followed immediately by a low-dose transmission CT which takes a couple of seconds. This is followed by rubidium injection followed by five to seven minute LISMO acquisition of the rest image. Because of the 76 second half-life of rubidium you can actually stress the patient right away after the rest acquisition is done and acquire the stress image along, inject the tracer and acquire the stress image at peak stress. And after that we typically also take another low-dose CT scan for attenuation correction for the stress images. In patients with suspected coronary artery disease we also add a coronary artery calcium score. If it's not done before you can do it either at the start of the study or at the end of the study. As you can see here this whole protocol takes about 18 to 20 minutes to finish and this leads to significantly improved efficiency and throughput of testing as well. Now let's discuss all the information that we get from a PET myocardial perfusion test and see how that performs with regards to diagnosing coronary artery disease risk stratifying patients who are at a higher risk for future adverse cardiovascular events and how we would help us guide our post-test management strategies. First we have relative perfusion assessment. This is similar to what you get with SPECT where we assess the extent and severity of reversible and fixed perfusion effect to get an estimate of ischemic and scar myocardium. All the properties that we discussed before actually do contribute to the higher diagnostic performance of PET using just relative perfusion alone. For example the high myocardial extraction fraction, better acquisition characteristics, and availability of ejection fraction wall motion abnormalities at peak stress as well as myocardial flow quantitation. All of these makes it more likely for us to identify disease in a patient where disease truly exists meaning it leads to a low false negative rate and contributes to the high sensitivity of PET. Similarly you know the high spatial resolution, the attenuation and scatter correction that we get with PET and the less GI interference that you get on a PET study. All of this makes it much more likely for us to confirm that the disease truly exists and with a low false positive rate leading to meaning it contributes to high specificity of a PET perfusion study. In the systematic review of 15 PET studies and 8 SPECT studies using obstructive disease on coronary angiogram as the reference standard, PET had a pool sensitivity of 90% and a pool specificity of 88% significantly greater than that of SPECT which was 85% sensitivity and specificity respectively. This was even seen in a head-to-head comparison in the same patient. In this study of about 208 patients with suspected coronary artery disease, PET had the best, the highest diagnostic accuracy at 85% compared to SPECT and coronary CTA performed in the same patient. Next we also, the next information we get is comparing the volumes at stress and rest and looking for transient ischemic dilatation. As we discussed before the PET images are acquired, the stress PET images are acquired at peak stress unlike SPECT where it is a 30 to 60 minute post stress image acquisition. Because of this the TID that you see on PET reflects true stress-induced cavity dilatation and not only subendocardial ischemia or post ischemic stunning that you see with SPECT. You can also expect as a result that it's more commonly seen on PET compared to SPECT studies. The upper limit of normal that we use, normal ratio that we use to diagnose TID is 1.13 to 1.15. This is from two studies that were published on this topic. And presence of TID actually contributes to increased specificity of PET. Other important differences of TID with PET is that you see TID in PET not only with multi-vessel disease but you can also see it in patients who have a significant single vessel obstructive disease as well. For example, in this patient where you see apparent transient ischemic dilatation visually on PET, the cath showed a significant proximal LAD, single vessel disease in the LAD. Not only that, similar to SPECT, PET is an adverse risk marker for long-term adverse cardiovascular prognosis, meaning it does predict patients who are at a higher risk of cardiac events in the future, but not only in those with abnormal perfusion. Actually, it is an adverse risk marker in patients with normal perfusion as well, unlike SPECT. LV function evaluation that we get on gated images with PET again is at peak stress. It's important to remember that. So we're much more likely to identify stress-induced wall motion abnormalities from ischemia as well as a drop in EF from severe ischemia at stress. Again, because of that, it is much more commonly seen on PET compared to SPECT. And all of this makes us more likely to identify disease if it is possible, meaning it increases the sensitivity of PET. Another important concept to remember with PET is this concept of LVEF reserve. LVEF reserve is the change in EF with stress. Normally, in normal patients without any significant disease, the EF on stress rises by about 5% or more. However, there is an inverse relationship between LVEF reserve and perfusion abnormality, meaning in patients who have severe ischemia who have very large perfusion defects, they have lower LVEF reserves. And LVEF reserve actually helps exclude a normal LVEF reserve of more than 5% has a high negative predictive value. And as you can see here, it can help us exclude severe three-vessel disease or left-main disease. And conversely, a blunted or a negative LVEF reserve where the stress EF falls compared to rest is very sensitive and helps us in identifying patients who have more extensive and severe coronary artery disease, especially if they are associated with an abnormal perfusion defect as well. But perhaps the most important information that is available from a PET imaging study is a myocardial blood flow quantitation. With PET, we can quantify myocardial blood flow at rest, stress, and also measure myocardial flow reserve, which is a ratio of stress myocardial blood flow to rest. FFR, which we measure invasively, uses pressure, a pressure drop across a lesion as a surrogate for flow and helps us identify significant focal epicardial stenosis However, myocardial flow reserve integrates the effect of significant disease across the entire coronary circulation, including the microcirculation and myocardial perfusion. PET is the most accurate non-invasive means to quantify myocardial flow as well as determine myocardial flow reserve. And you can measure it in the same study that you use to look at the relative perfusion images using dynamic LISMOD data without the need for any additional radiation. Myocardial flow reserve on PET, similar to the EF, also helps us improve the diagnosis of coronary artery disease, especially multivessel disease. In this study of 120 patients who underwent dipredermal rubidium PET, we noticed that below a flow reserve of two, the probability of having three-vessel disease increases with lower myocardial flow reserve. Not only that, it significantly added to the diagnostic performance of just perfusion alone. And 88% of the patients who had severe three-vessel disease in this study, all had reduced flow of hyperemic myocardial blood flow as well as reduced global myocardial flow reserve. In this study, we see that a normal myocardial flow reserve defined as a myocardial flow reserve of two or more has a very high negative predictive value and helps us exclude severe left main and multivessel disease in patients with normal or mildly abnormal perfusion. So when do myocardial flow measurements offer the most value? As we saw in the last two slides, it helps us rule out significant high-risk disease. It is also useful in patients with suspected microvascular dysfunction. These are more likely to be women and those with cardiometabolic disease risk factors such as those with diabetes, chronic kidney disease, obesity, et cetera. It's also useful in patients with suspected left main or multivessel disease where you suspect balanced ischemia and suspect that the perfusion images itself might not capture the severity and the extent of the true disease. And it's also helpful in patients who have known disease such as those with known chronic total occlusions where you want to know the significance of the obstructive lesion. It also helps us provide certainty regarding single vessel involvement versus multivessel involvement. And also it helps us assess the response to vasodilator and assess if it is actually a true stress study or if the patient had caffeine or other circulating antagonists. Now let's look at practically how does myocardial flow quantitation helps us with diagnosis of coronary artery disease phenotypes. In this patient, a 70-year-old female who had hypertension, diabetes, obesity presented with symptoms. We see the perfusion image here and all of us can agree that it looks completely normal. Now, whether it's a truly normal study or whether this patient has a left main disease or balanced ischemia from left main or multivessel disease or whether she has microvascular dysfunction or whether the patient did not respond to vasodilator at all, we can't say just looking at relative perfusion alone. And this is where myocardial flow data helps us differentiate between these various phenotypes. If the blood flow augmentation is normal regionally as well as globally, you can see here that the flow reserve is normal more than to its 3.5. This suggests that this is truly a normal study, that the patient had a good vasodilator response and she's at a low risk for future adverse cardiovascular events as well. Or the flow data can look like this, where you see a regional decrease in the flow in the left circumflex territory. However, globally, the flow reserve is normal. If the patient had normal perfusion in this case, the myocardial flow data has now helped us uncover single vessel obstructive disease. And if the patient had a left circumflex disease perfusion defect, it helps confirm that the disease is limited to that territory and there's no significant disease in the other coronary arteries. Or you can see a pattern where flow is diffusely reduced globally at stress and the flow reserve is blunted in all coronary territories. This pattern is seen in patients with microvascular dysfunction. Or you may again see the same pattern of severe reduction in myocardial flow, in peak myocardial flow in all coronary arteries. Along with a significantly reduced myocardial flow reserve, especially if it is 1.5 or less, you should be really concerned about multivessel disease, such as this patient who had significant three vessel disease along with left main disease. You can also see it in patients with severe microvascular dysfunction. However, such severe reductions typically require some form of coronary anatomy evaluation before just medical management. And if the flow does not augment at all with stress or the flow reserve is around one, that should raise the end of the patient as a normal perfusion, that should raise suspicion for a non-response to vasodilator or presence of caffeine or other circulating antagonists. The other useful information that we get from a PET study is looking at coronary artery calcium. You can also identify this on just the low dose CT maps that we do for attenuation correction, or you can calculate the score from a coronary artery calcium score that's added to the study. But that is not the only thing that we do. We also look at the coronary artery calcium score that's added to the study, but that helps us identify calcified atherosclerosis and not only would help with patient management with regards to initiation of preventive therapies in a patient who does not have any known coronary artery disease to begin with, but it also helps us with further prognostication to the other markers that we get with PET. All of the markers that we saw leads to, and the characteristics of PET contribute to improved diagnostic confidence with PET compared to SPECT as well. You see in this study, experienced readers were more likely to call a PET study as definitely normal or definitely abnormal versus SPECT, which was more likely to be called probably normal or abnormal or equivocal. And this improved diagnostic confidence with PET also translate to more appropriate downstream patient management compared to SPECT. Let's look at this study from our center in Kansas City where this was a randomized study of about 322 patients with known coronary artery disease who were symptomatic. These were randomized to either PET or SPECT and followed for 12 months for downstream resource utilization. And we see here that patients who have a high risk PET finding were actually appropriately, so more likely to be referred to cath and revascularization compared to those who had a high risk SPECT study. Also in patients who had a low risk PET were less likely to be referred for cath or revascularization compared to those with low risk SPECT. Now let's look at, we've talked extensively about how PET can help us in diagnosing not only significant flow-limiting obstructive epicardial disease, but also microvascular disease. Now let's look at how these PET markers help us in prognosticating patients and risk stratifying those who are at a higher risk for future adverse cardiovascular events. In this study of about 7,000 patients across four centers, similar to what has been shown with PET, we see that addition of a relative perfusion marker of a percent of abnormal myocardium, this is a combination of reversible and fixed perfusion defect. When you add this, when this information is added to just pretest clinical risk factor, it does provide improved, it leads to improved risk reclassification for long-term cardiac death in these patients compared to just clinical risk factors alone. Like we know with SPECT, REST-EF that we get on PET can also help us further risk stratify patients beyond perfusion into identifying those patients, those with lower ejection fractions at REST who are at a higher risk for future adverse cardiovascular events. Not only that, the LVEF result, the change in EF with stress that we get from PET also provides incremental prognostic value beyond just REST function as well as perfusion. You see here in this study of about 1,400 patients that those who had a negative LVEF result, meaning the EF dropped with stress, had a much higher rate of cardiovascular events and all-cause death compared to those who had a normal or a preserved LVEF result. The same applies to myocardial flow result. In this study of around more than 12,000 patients who underwent rubidium PET myocardial perfusion imaging at our center in Kansas City, we noticed that perfusion does help risk stratify the population and identify patients who are at a higher risk for future events. However, within each strata of perfusion, even in those with completely normal perfusion and no ischemia or those with severe ischemia, those who had a reduced myocardial flow result had a significantly higher risk for cardiac death as well as long-term death compared to those with preserved myocardial flow result. And every 0.1 unit of decrease in myocardial flow result was associated with a 9% increase in the hazards of cardiovascular death as well as all-cause death. This was also seen in other series from multiple other risk factors, sorry, multiple other centers, including thousands of patients, where again, everything consistently shows the same finding that beyond perfusion, even in patients who had normal perfusion or those with severely abnormal perfusion, those with lower coronary flow result or myocardial flow result have a higher risk for future adverse cardiovascular events, including non-fatal MI, heart failure, as well as all-cause death. Now, let's look at how these PET findings can help us guide our post-test management decisions. Can it help us identify those patients who may benefit from revascularization or not? We looked at this in 16,000 patients from our center in Kansas City, where we noticed that, so this is, let me actually explain this figure. This figure is a plot of the hazard ratio of death with early revascularization compared to medical therapy. So everything that falls below one actually means lower death with revascularization compared to medical therapy. And everything that goes above one means higher death with revascularization compared to medical therapy. So you can see here that as the percent ischemia increases, especially in patients with more than 10% ischemia on their PET studies, these are the patients who have a benefit, meaning they have a lower death rate with revascularization compared to medical therapy. And another important difference here compared to the SPECT data that we know is that this threshold where we start seeing the benefit might be lower with PET compared to that we see with SPECT. So here we see that you start seeing the benefit between five to 10%, whereas with SPECT, we know that the ischemic equipoise threshold is between 12 to 15%. We extended this beyond just cardiovascular events and death. We looked at symptom and health status as an endpoint, and we saw a similar result where patients who had more than 10% ischemia on their PET perfusion study had a higher chance of improving their angina as well as their quality of life and overall health status when they underwent revascularization compared to medical therapy. So these are odds ratio with revascularization compared to medical therapy, and you see that they have more than a threefold higher odds of improvement in their angina at 12 months and their health status at 12 months when they undergo revascularization compared to medical therapy. And again, it's important to know that all of these patients in this small study had their known coronary artery disease and were on good medical therapy. So this benefit was seen on top of that. Now let's take it one step further. We noticed in this study of more than 12,000 patients at our center that myocardial fluorescence actually adds and further helps us identify patients who may benefit from revascularization beyond just relative perfusion assessment alone. You can see that the patients who have a reduced global myocardial fluorescence actually have a survival benefit when they undergo revascularization after a PET study compared to just being medically treated. And that we saw this difference with both type of revascularization, whether it be PCI or CABG. And that combining the markers of perfusion as well as myocardial fluorescence can actually help us better select patients who will benefit from revascularization. As you see here, whereas there was a trend. So these forest plot actually plots the hazard ratio of death with revascularization compared to myocardial fluorescence. Compared to medical therapy. So again, same thing, less than one benefit, more than one harm. And you can see here that those patients who had significant ischemia, only those who had reduced global fluorescence along with that were the ones who actually had a survival benefit when they had revascularization after their PET studies compared to medical management. And conversely, even if you have a perfusion defect, you may not get a benefit with revascularization if your global flow reserve is normal. So to summarize this talk, PET has numerous advantages over other studies such as SPECT, including much better resolution, improved image quality, lower effective radiation dose to the patient and better efficiency of testing. Unlike SPECT, the stress images at PET are taken at, are acquired at true peak stress. And that leads to these metrics such as transient ischemic dilatation and NVF reserve that offer, add to just perfusion images as far as diagnosis and prognosis goes. All of this PET has much superior diagnostic accuracy, not only to diagnose obstructive epicardial disease, but also diagnose other phenotypes of coronary artery disease such as microvascular dysfunction. And then myocardial flow quantitation and flow reserve on PET also helps, not only helps in diagnosing disease, but also helps further risk stratify patients who are at a higher risk for adverse cardiovascular events in the future. And in the end, PET has a promising role in helping us identify or select patients who will benefit from revascularization compared to just medical treatment alone. With this, I end my talk. If you have any questions, please feel free to reach out to me. I'll be happy to answer them. Thank you.

cardiac PET myocardial perfusion imaging

PET imaging

positron emitting radio tracers

SPECT scanners

pharmacologic stress testing

image quality

risk stratification

myocardial blood flow

gamma rays

spatial resolution

myocardial flow quantitation

diagnostic accuracy

SPECT