Hello, I'm Sharmila Dorbalam from Brigham and Women's Hospital in Boston. I'll be speaking to you about imaging protocols and tips for exam interpretation. These are my disclosures. During this talk, I'll discuss how to implement PET myocardial perfusion imaging protocols, how to optimize the imaging protocols, and how to systematically evaluate PET myocardial perfusion images. Preparation for PET MPI is standard and same as for SPECT MPI. Patients come in NPO and have no caffeine for at least 12 to 24 hours. Withholding of medication, if possible, is recommended for patients with no previous known coronary artery disease. We have three approved radiotracers for PET myocardial perfusion imaging. Rubidium 82, a short-acting radiotracer which requires an on-site generator, N13 pneumonia with a 10-minute half-life requires a cyclotron on-site or very nearby, and F18 floperidase, which is recently approved, has a long half-life of 109 minutes and can be delivered as unit doses from a remote cyclotron. Therefore, no on-site generator or cyclotron is required for this particular tracer. In terms of stress agents, we could use all vasodilator agents, dobutamine, and we could also use exercise for PET myocardial perfusion imaging. So what about stress protocols for PET? The most commonly used protocol is trigadenosone vasodilator stress. Trigadenosone is administered as 0.4 micrograms per 5 mL injection over 10 to 20 seconds, followed by a saline flush. The radiotracer is administered approximately 55 seconds later. For PET imaging, we recommend that the PET scanner be started first, followed by administration of radiopharmaceutical. All three radiotracers can be used for trigadenosone stress. Dipyridamol and adenosine are weight-based infusions administered over 4 minutes. Dipyridamol, a standard dose, is given for 4 minutes with a radiotracer injection between 7 and 8 minutes. The PET scanner again is started before the radiotracer is administered. With adenosine, it's a 6-minute infusion with administration of radiotracer at the midpoint, around 3 minutes. Again start the PET scanner first and then give the radiotracer, and this becomes very important for dynamic imaging, which is essential for quantifying myocardial blood flow. In patients with contraindications to vasodilator, dobutamine stress could be used. Standard protocols apply here, starting with 10 mikes per kg per minute, going to 20, 30, 40, and use of atropine if needed to get to 85% of age predicted maximum. To start the PET scanner, inject radiotracer, continue dobutamine for 1 additional minute to maintain peak hyperemia during the PET acquisition. Exercise PET is also possible with N13 ammonia and fluorpyridase. You could use either treadmill or bicycle exercise. Protocols are standard. We inject the radiotracer at peak exercise, continue exercise for 1 to 2 minutes to maintain hyperemia, and then stop exercise. After termination of exercise, there's a brief wait period after which the patient is positioned in the PET scanner. This wait period varies between radiotracers, and with ammonia it's 2 to 3 minutes. With fluorpyridase it's 5 to 25 minutes. So why do we need to wait? It's primarily because the patient is exercising heavily with breathing and motion following exercise. So if you wait a few minutes, then breathing settles down and the PET images will be less artifacts in the PET images because of breathing motion. With fluorpyridase you have the opportunity to wait even longer because of the long half-life. Remember when you use exercise with positron emission tomography, myocardial blood flow cannot be evaluated because you cannot obtain images at the time of radiotracer injection. So you don't have images or input function for myocardial blood flow. What about imaging protocols? Myocardial perfusion PET includes three common steps. The first step is a scout image to position the scanner to find the heart. The second image is a CT image. This is for a transmission scan for attenuation correction. After this is completed, radiotracer is administered and an emission image is obtained. Following this, most commonly pharmacologic stress is used. Radiotracer is administered based on the pharmacologic stress as we discussed previously, followed by a stress myocardial perfusion study. If the patient remains in the PET gantry for the stress and rest without motion in between, all you need is one localizer and one CT scan for attenuation correction. However, if the patient completes the rest and is moved out of the PET gantry, remember you may have to reposition the patient and repeat the CT scan for attenuation correction. Some more details about long-acting radiotracers. So if you use N13 ammonia with a 10-minute half-life, you could perform a protocol with equal dose. You administer N13, get your rest pictures. You need to wait 50 minutes between the two equal doses to allow for clearance of the resting activity in the myocardium. Another option is to use a low-dose high-dose protocol, where you administer a low dose of N13 ammonia, wait 10 minutes and administer three times the dose and this way you can the entire study in a shorter time period. With fluid periodonts, the minimum time between the low-dose rest and high-dose stress dose is 30 minutes when using pharmacologic stress and 60 minutes when using exercise stress. How do we interpret these PET images? As with SPECT myocardial perfusion imaging, you have perfusion images, you have polar plots, you have gated myocardial perfusion images. Remember with PET imaging, you gate the study both at rest and at stress. If the ejection fraction decreases from rest to peak stress, even with vasodilator stress, that indicates a high-risk feature and that has been associated with multivessel obstructive coronary disease or left main disease. In addition, we also calculate myocardial blood flow using the dynamic images. And of course, for hybrid scanners, we can look at coronary artery calcification either on the low-dose non-gated CT, which is acquired for attenuation correction, or on the calcium score CT, which we perform in all patients who have no previous documented history of coronary artery disease. So basically, interpretation goes through quality control of perfusion images, polar plots, gated images, myocardial blood flow images, as well as the calcium score images. I'll focus today on perfusion imaging interpretation. And these are the steps that we'll follow in interpreting these scans. We look for adequacy of the attenuation correction, evaluate for motion, lung uptake, excess blood pool activity, quantify myocardial blood flow, and then of course, lastly, ensure that the patient has had an adequate hyperemic response to stress. So here's an example of a patient. Stress and rest images are shown here, rubidium 82. Here are the polar plots and all these images show an anterolateral perfusion defect on the stress that is completely reversible at rest. You can see that here, here as well. Polar plots confirm the same. On the right side, you'll see fusion images. Gray scale is a CT scan or the transmission scan, and the color scale is the emission image overlaid on the CT scan. And what you'll notice is the emission image is not adequately registered with the CT and is spilling over into the lung fields. Because of this, there is an artifact and there's reduced counts in the anterolateral world with an apparent reversible anterolateral defect. So when you don't see adequate registration, these images need to be reconstructed after registration of the emission and transmission images, which is what this slide shows. Once we correct the registration of the images, you can see that there's no longer an anterolateral reversible defect and the polar plots also show significant improvement compared to the baseline study. Next we look for motion artifacts. Here's an example of a stress-rest study on a patient. The resting images look high quality. However, when you look at the stress images, they're very blurry and you can see the shape of the ventricle is also off and the myocardium looks much thicker, especially focused on the anterior wall here compared to the rest. So why is this? So all of this suggests that there could have been patient motion during the stress image acquisition. Unlike SPECT imaging, with PET imaging detection of motion can be challenging. We could look at the dynamic frames and play them in a cine loop to detect motion. Here you see stress images on the top, rest images on the bottom, and you'll focus on the stress. You can see the heart is moving down and then moving up later on. Down here and then moving up. So significant patient motion in the stress images. Now focus on the rest, you see no significant motion. It's a pretty steady. So basically this patient experienced motion during the stress image acquisition which accounts for these odd images and the odd shapes. Some softwares will allow for motion correction for the dynamic frames, specifically for blood flow acquisition. And then you can see motion even on this graphical display shown in millimeters. What could you do with motion? There are ways to correct for motion. They're not optimal, but we could do that. So on the top is the first image that we reviewed, motion. On the bottom is a motion corrected. So what do you do here? You take the multiple frames from the dynamic images and discard the frames which are impacted by motion and sum the rest of the images which are good quality. When you do that, you can see excellent quality stress and rest images. So motion is difficult to detect. It's important to pay careful attention during the image acquisition. We can reframe the emission images, and in this case it seemed to have worked. But most important, remember, do not review the non-attenuation corrected images because they could have artifacts on them with PET. Rarely, if none of this works, you may need to repeat the study, both emission and transmission. Increased lung uptake is another potential artifact that you could see, particularly with N13 ammonia. Here's an example of stress and rest N13 ammonia perfusion images. The polar plots show an apparent reversibility in the lateral wall. When you look at the perfusion images though, there's that severe lateral wall defect, both at stress and at rest, but there's this excess activity in the lung on the rest images. And this is what is accounting for the reversibility on the polar plots. When you look at the fusion images, you can see there's a lot of lung uptake here. So why is lung uptake increased in this patient? This could be seen in patients with a low ejection fraction, and this patient had an EF of 25%. It could also be seen with N13 ammonia sometimes in smokers. So please remember that when you see this, this could be an artifact related to N13 ammonia. Another normal variant with N13 ammonia perfusion images is the so-called fixed basal lateral wall defect. So focus on your stress and rest images. This is this lateral wall perfusion defect, which is seen both at stress and rest. It's also seen on the polar plots here. So when you see this kind of a defect, and if the gated study is normal, which was the case here, then this could be a normal variant with N13 ammonium. In patients who have paradoxical septal motion, either because of a left bundle or prior bypass or cardiac surgery, this seems to be somehow exaggerated. And in this case, you can see that there's almost a double shadow, which is also related to partly to that. One challenge with rubidium-82 could be excess blood pool activity. So here are resting images of a patient whose ejection fraction was 25%. You can see the signal-to-noise ratio is somewhat limited. We reconstructed the images a second time, and here the images are much improved quality and the blood pool is better. So what did we do? So the standard reconstruction for rubidium-82, the pre-scan delay, the delay between injection to the time where you start acquiring these images is 70 to 90 seconds with rubidium-82 and normal EF. So if you do this, and if you have a lot of blood pool activity, you could extend the pre-scan delay. And in this case, we extended to 120 seconds, and therefore, you can see excellent clearance of blood pool. So in patients who have low ejection fraction, look for blood pool activity that's present, extend the pre-scan delay. What about myocardial blood flow? This is one of the added values of PET myocardial perfusion imaging, and in fact, it's a key reason why one may consider PET myocardial perfusion imaging. Here's an example of a patient who has an inferior wall perfusion defect that was reversible. When we look at the blood flow value stress, you see the RCA territory flow values are very low compared to the circumflex and LAD. However, when you look at the flow reserve, it's all normal, more than two being normal. And here are the time activity curves, time here, and then activity on the vertical axis, and pay attention to this green, which is the left ventricular curve during stress, all right? And here are the myocardial counts. So I'll show you now myocardial blood flow computed on this patient a second time. And here you'll see that the blood flow in the RCA and circumflex are even lower, and here the flow reserves are all abnormal. Global myocardial flow reserve is severely abnormal, and so is it regionally. So what's the difference between the first and the second reconstruction? By optimizing the position of the input function, so I moved around the input function, I was able to maximize the input function. Here you can see the counts are much higher for the input function compared to the initial reconstruction. So basically, it's important for us to optimize or maximize the input function so that you get the lowest possible blood flow values. Artifactually, it's easy to overestimate blood flow, but underestimation of blood flow is typically not a challenge. Therefore, whenever you can get the lowest possible blood flows, those are the ones you should be reporting, and keep moving your input function until you get the maximal peak activity, as in this case. So here now, this patient has severely reduced myocardial flow reserve in all three territories, and of course, much worse in the right coronary and circumflex compared to the LAD, a very different interpretation compared to the initial values. The last thing I want to speak about is adequacy of hyperemic response. So of course, if you have a scan with high hyperemic blood flow and high flow reserve, then you already know that's an excellent quality interpretable study. With exercise or dobutamine stress, if you have the patient achieve more than 85 percentage predicted maximal heart rate, that is also adequate hyperemia. With vasodilators stress, we do look for something called splenic switch-off. If splenic switch-off is present, then that is adequate hyperemia. If there's no splenic switch-off, then we may need to consider repeat study. Here's an example of a patient. You see rest images on the bottom, stress on the top. The arrows point to the spleen. You see there's no change at all from rest to stress and a physiological response with stress would be decreased splenic activity with stress. So here the flow values are very low and the flow reserve is hovering around 1. So what do you do when you see a patient like this? It is extremely challenging to call these flows and flow reserves as severely abnormal because you see no splenic response in this patient. So what would you do? You could repeat the study. In this patient, we did repeat the study. Here look at the spleen now in the repeat study. Excellent splenic switch-off and look at the myocardial flow at stress. Blood flow values are substantially increased compared to rest and your flow reserve is normal. So this is how you would look for hyperemic response and adequacy of stress before calling very low flow values as abnormal. So to summarize pet myocardial perfusion imaging, I have shared with you some specifics about how to implement stress protocols, how to optimize imaging protocols, and how to systematically interpret images. We discussed the three possible radio tracers for pet perfusion imaging that are used clinically these days, 82 rubidium, N13 ammonia, and recently approved 18F fluorpyridase. All available stress agents could be used for pet MPI. Most commonly used regged adenosine, exercise stress is feasible with ammonia or fluorpyridase. And if there are contraindications to vasodilate or dobutamine could be used. Interpretation of stress perfusion images is similar to SPECT with some additional points to keep in mind. Look for adequacy of attenuation correction, look for patient motion, lung uptake, and blood pool activity depending on the radio tracer. Carefully evaluate the myocardial blood flow values and ensure that there is adequate hyperemic response before calling the low blood flow values as severely abnormal. With that, thank you very much for your attention.

PET myocardial perfusion imaging

radiotracers

stress protocols

image interpretation

motion artifacts

quality control

PET myocardial perfusion

imaging protocols

myocardial blood flow

image artifacts