Hello, my name is Moaz Milah, I'm the Director of Cardiovascular PET at the Houston Methodist DeBakey Heart and Vascular Center in Houston, Texas. And I'll be talking to you about my cardioperfusion imaging protocols using PET imaging. These are my disclosures. So the most important step when starting to image a patient is patient preparation. We want the patient to be well prepared so that they can undergo the test with limited artifacts and to pass them through the test without any limitations. So first of all, there are some general preparation instructions that we use for everyone who come for my cardioperfusion imaging. We prefer them to be MPO for four to six hours. We want them not to have caffeine, especially that we are using vasodilator stress, so we prefer them to be for 18 to 24 hours without caffeine, keeping in mind that you may require them to be longer for some patients if they are used to use a lot of caffeine. And finally, you prefer that they don't use the offering containing medications for at least 48 hours. They could antagonize the effect of vasodilator stress, but also this is the antidote medication that we use during stress testing. When the patient comes to the lab, the nurses assess them and the physician or the person who's going to perform the stress test assess them for contraindications for the stress test. And finally, once that's determined, we put them on the table of the PET machine after inserting an IV, preferably in the right antecubital or in the right arm, as long as it's a good access. The patient should be in the supine position and the arm should be up and out of the camera field of view. Obviously, some patients might have arthritis and may not be able to keep their arms up. And if you have a PET CT, this is less of an issue because of attenuation correction. In patients with severe arthritis, images can be obtained with the arms on both sides, but it is very important to have the arms on both sides on both stress and rest imaging. And finally, we want to keep the patient in the same position as much as possible. So the basic elements of the protocol include doing a scout image, then we're going to do a transmission scan, and most commonly we are using CT attenuation correction nowadays. And finally, we'll do an emission scan, which is a nuclear scan, and we've used them between the nuclear and the CT images. We repeat these steps in stress and rest. You may not need to repeat the scout unless the patient falls out, but you may need to repeat the transmission scan depending on your lab protocol. The acquisitions also can be done as EKG gated. So like we do in SPECT, we acquire the images and then they are binned according to the EK timing in the EK and the cardiac cycle based on the EKG. You can do eight bins or 16 bins. That was one way of acquiring images. And then these are the images as you see them here. Once we play them, this is a 16 bin, and this will allow you better temporal resolution, allowing for better ejection fraction assessment in some cases. Or you can acquire the data as dynamics. Now we're not acquiring it in relation to the EKG, but in relation to time. So as you see here, we see rubidium and this is a rubidium study. We see rubidium coming into the right ventricle and after that, it's going to go to the left ventricle and then into the myocardium. So right ventricle and then left ventricle and the myocardium. So the acquisition parameter here is primarily time. So irrespective whether this is happening in systole or in diastole. And then we bin the data according to time to allow us to do a myocardial blood flow. Now most scans right now allow you to do both acquisitions at the same time using LISMOD acquisition. And this is how we bin the data for dynamics. So we take initially the right ventricular data, then the left ventricular data, then the myocardium and it's reconstructed as you see here in time activity curves. In LISMOD acquisition, we acquire the data irrespective where they happen in relation to the EKG or time. And then after the acquisition, we go back since we have the EKG data and bin it into EKG and to gated data or bin it according to time and get the information in relation to EKG and time and then have a gated series, a static series, which is the sum of the gated series. And then finally the dynamic, which is in relation of time. One thing to keep in mind that in the static or the gated, we don't want the radiotracer to be in the blood pool. We want it to be outside the blood pool. So we want primarily to acquire the images or have the images only in the time where we have a clear blood pool and most of the rubidium or the ammonia is in the myocardium. And for that, we have to take out and not include the early phase. So this is a time which is called pre-scan delay, which we add here and usually it's about two minutes in normal hearts and normal LV function. And this is important to allow the blood pool to clear and now we only sum the images that are acquired when the LV cavity is clear and most of the radiotracer is in the myocardium. You could do the same thing for the gated image. However, for the dynamic image, we want the data from the beginning and this is extremely important that we start our image acquisition as soon as we inject or preferably a few seconds before you inject to allow you to gather the information on the right ventricular phase, the left ventricular phase and the myocardium. If you miss this period, then you will not be able to acquire dynamic data and calculate myocardial blood flow. So I'm going to show you now protocols related to the most commonly used stress myocardial perfusion imaging radiopharmaceuticals. I will focus primarily on rubidium and ammonia. These are the only two agents that are FDA approved in the United States. For rubidium-82, the parent molecule strontium-82, you usually get a generator every 42 days and the half-life is about 75 seconds and the dosage differs based on the type of the machine you have. So if you have a 2D machine, and it's very important to know the type of machine you have. This is not by vendor, but rather by the characteristics of the machine. If you have a 2D machine, then you need to do 40 to 60 millicuries of rubidium to be able to get good image quality. If you have a 3D machine, you do 25 to 30 millicuries depending on the patient body surface area or based on the patient weight or BMI. And finally, if you are trying to do, if you have one of the newer generation machines like a digital pet, then you can acquire with ultra-low doses of 10 to 20 millicuries. And it's usually injected as a rapid bolus infusion in 30 seconds. This is a protocol which is probably most commonly performed in nuclear labs, in pet labs. So what we do, and here I put the doses for 3D machines. So you have to adjust that according to the type of machine that you have. So we start with the Scout as we showed in the beginning. We do the CT attenuation correction. We inject the radio tracer and then start immediately the list mode acquisition. If you have that capability on your machine, which is most systems now that are in the install base have that ability to do so. And then later on, you unlist your data and get the information of the gated and dynamic datasets. After that, you have to wait usually about 10 minutes between the two injections. So you have about seven minutes of list mode, one minute to get the patient ready, or one to two minutes. Then you inject the Rigadenosone, wait about a minute or so, and then inject Rubidium again. Acquire and list mode acquisition the stress data. You may or may not repeat the CT attenuation correction, depends on your lab protocol and whether the patient moved or not. And you may or may not perform a calcium score if your lab allows you to go ahead and acquire the calcium score for quantification of calcium atherosclerotic burden in these patients. So this protocol usually takes about 25 minutes. It's extremely important that you use the Rubidium dose that is more appropriate to the type of system that you have in your lab. And as I said, some labs may not do a second CT attenuation correction. That depends on your patients, but also on your local lab protocols. If for whatever reason, the patient cannot get Rigadenosone due to a contraindication or history of seizure, for example, then the patient, then you can use Adenosine or Dubitamine. The protocol is not much different other than it's preferred to have two IV lines because after you inject Rubidium, you still have to continue the Vasodilator or Dubitamine for a few minutes after two or three minutes. And during that time, you don't want to push a lot of Adenosine in the same IV. So it's very important that you go ahead and prefer to have two IV lines. One for the Vasodilator or the stress agent, and the other one is for the injection of Rubidium. Unless you have one of those newer three-way injection systems that will not push a lot of data, a lot of Adenosine in the patient's arm. Other than that, the protocols are the same. As you can imagine, these protocols are longer, especially if you're using Dubitamine. If the patient has contraindication to Vasodilator stress. And for Dipyridamol, it's very similar. The only difference that you do here is that you want to inject your stress dose at peak hyperemia. So with the injection at the time of peak hyperemia, with Dipyridamol, that happens around seven minutes after the initiation of the injection. So you stop. So once you complete your four-minute injection of Dipyridamol, you wait three minutes for maximal hyperemia to happen, then you inject your Rubidium and follow it up by the rest of the steps as you have in other protocols. There are a couple of practical considerations to do in regards to stressing the patient and the new room design. It's very important that once you inject the Rubidium for both the rest and stress, that the staff will be asked to leave the room for the first few minutes because otherwise they will have high cumulative exposure to the Rubidium doses or radiation exposure. A very important aspect is that also a lead apron is not helpful in shielding the 511 kV photons. So by having a lead apron, you're not protecting the staff or any family member who may be with the patient. Doses for Rubidium, at least we are using BMI-based doses. Some labs are using weight-based dosing, but it's very important to keep in mind that patients with large BMIs will benefit from higher doses. And 3D imaging in general requires less dosage. If you give high doses with 3D, then you could saturate the crystals and will impact the myocardial blood flow measurements. So we mentioned this, it's very important that you start imaging at the same time of the injection and then you later on add a pre-scan delay for the static and gated reconstruction. You do not add a pre-scan delay for the dynamic reconstruction. The duration of this pre-scan delay depends on the ejection fraction of the patient. So if the patient has a low ejection fraction, their pre-scan delay may be a little longer than someone with normal ejection fraction. So here you look at this patient. There's a lot of blood pool activity in this image, and this patient will benefit from longer pre-scan delay. So here with 120 seconds or two minutes, there's still a lot of blood pool activity, and you may need to repeat it at 150 seconds or 180 seconds sometimes, as long as you have good counts in the myocardium and good image quality. So it's a trade-off between clearing the blood pool and having enough rubidium in the myocardium to be able to give accurate diagnosis. Here's another example, even in patients with normal ejection fraction, but they have a low flow state, there is still some cavity rubidium, and once we increase that adding 30 second delay, now you can see the image quality and the edge detection or delineation is much better in here. There are fewer tips that it is very important for your technologist to keep in mind. One of them is the issue of misregistration. So if the patient moved between the emission and transmission scan, then you may induce a defect which may not be true, as you see here, and when you look at the fused images, you see that the heart is partly over the lungs, and that's why you have an anterolateral defect. It's extremely important for the techs to recognize that, and they might have a hint if they see the patient moving, and this happens most commonly during stress, especially as the patients experience side effects from the vasodilator that is being injected. And once that is corrected nowadays, all vendors have tools that will allow you to reconstruct it, but this has to be done on the council. It's not going to be done at the level of the interpreting software, then once you fix it on the council and reconstruct your data, this defect is completely gone away. Another important aspect, I emphasize that you start your image acquisition at the same time as you do the injection, or preferably a few seconds. So you start the imaging, then the injection. Here's a patient where you see one peak rather than two, and you see you're missing the other peak when you run the motion, the movie here, you can see that at rest, when we start, there is no rubidium, then it comes to the right ventricle as we would like, but with stress, this one was started a little bit after, so they missed the input function. And obviously this patient has, you cannot compute the peak myocardial blood flow. So more on the safety margin, either you start them at the same time, or maybe one second earlier you start imaging, then you inject to make sure that both of them are starting at the same duration and you don't miss the peak of the input function. And finally, it's very important to work with the patient, so they limit their motion. Motion artifacts are common with PET, especially if the patient moves, because we're imaging them during maximal hyperemia as they develop some of the side effects from the vasodilator agent, so it is extremely important to limit their motion, otherwise you may have some uninterpretable studies. As you see here, this patient has significant motion in the inferior wall with a defect which is likely not true, and this was proven by angiography to be not true in this case. Few words about ammonia, it's cyclotron produced, the half-life is 9.8 minutes, so it allows you for easier dosing in terms of if you're a little bit far from the cyclotron. Again, the dose depends on the system you're using, so if you're using a 3D system, you may need lower doses, higher doses for the 2D systems, and the digital will require the lowest doses, and it's possible to combine with exercise, although there are not too many labs that are performing it routinely with exercise. The only drawback is that if you are injecting the same dose at stress and rest, then you will need some time in between the stress and rest to allow for the decay of ammonia in between the two doses, and usually that's estimated to be about 30 to 50 minutes between stress and stress to allow for imaging. And this is the protocol which is very similar to what you see, what you just saw with rubidium, so you do a CT attenuation correction, you inject the radiotracer at rest, do your acquisition, have the wait until the rubidium, until the ammonia that's injected decays, do another CT attenuation correction, or skip that depending on your lab protocol, do pharmacologic stress, inject at peak hyperemia, acquire images for 7 to 10 minutes, again depending on the system you have, and then you're done. There are some newer protocols using the newer digital PETs, so if you have a digital PET, then you may inject a lower dose here, an ultra-lower dose, then you don't need to wait this time, and you can just inject double or triple the dose for stress. So if you inject 4 millicuries at, or 5 millicuries at rest, you can still do a 15 millicuries at stress, and you'll still be fine in terms of the shine-through effects from the rest dose. If there are some sites who are doing exercise ammonia primarily for research, again the same concept applies here, but the idea is that you want to inject ammonia while the patient is exercising, have them exercise for about a few seconds up to a minute after, then you have to get them on the table for image acquisition within 4 minutes maximum to acquire your images and allow for image acquisition for the stress protocol. So I hope I showed you the protocols that are needed for a successful start of a PET-myocardial perfusion imaging. Thanks for your attention.

Moaz Milah

Cardiovascular PET

imaging protocols

patient preparation

PET machine

myocardial perfusion imaging