Welcome everybody, it's an honor to be presenting here in the PET curriculum. My name is Aldo Escanone, I do advanced cardiac imaging at the Brigham. And today we'll be discussing and talking a little bit about cardiovascular PET tracer, some of the basic concept, and we're gonna get a little more in depth in terms of perfusion tracers. I have nothing to disclose. So I guess to kind of get things going, the goal of this talk is to gain knowledge on the physics and physiology and unique characteristic of the PET flow tracers, but we're gonna do it within a practical point of view and how we can use that in the interpretation of images. Learning subjects of today's talk is gonna be to define what constitute a radio tracer and a perfusion radio tracer. We're gonna describe the physics, a little bit of biology and physiology of the tracers, and we're gonna compare the strengths and the weaknesses of the PET flow tracers that we have currently available. So the first question that one has to ask when we talk about tracers is what's a radio tracer? And a radio tracer basically is a chemical compound that has unstable atoms. And by virtue of the radioactive decay of these unstable atoms, it can be used to explore processes or chemical reactions. And these unstable atoms is what we call a radionuclide, and a radionuclide is not more than just an atom that has an excess of energy. And because that excess of energy makes this atom unstable and is prone to decay to try to reach the stability. When we talk about PET tracers, there's kind of four big categories as we see here. We can talk about tracers that describe perfusion, and that's kind of the category we're gonna make an emphasis today. And as an example, we have ammonia, rubidium, water, or fluid pyridase. We also have tracers that describe and explore metabolism like C11 acetate or palmitate. We have tracers for inflammation, as we all know with the FDG, and even tracers that are under investigation to analyze amyloid plaque. And as I said, today we're gonna be discussing perfusion tracer. And basically the characteristic of perfusion tracers is that they accumulate in target tissues in proportion to the blood supply. So that becomes very handy when we're trying to analyze the blood supply to the heart and do macular perfusion imaging. So before we dive in into the specifics of the tracer, I think it's important to do a little refresh. And I'm pretty sure that this was already kind of described in greater depth in prior modules. But just to kind of, as a mode of refresher, we know that PET-rated tracers decay by emitting a positron from the nucleus. And we have in this cartoon, we can describe how that process work. In our atom that is unstable with an excess of energy, in the case of PET tracers, they have an excess of protons. And the way that the atom reaches or try to reach stability is by switching or converting one proton, as we see here, into a neutron. And that process is accompanied by the release of energy and the release of a positron here in green. That positron travels until it reaches its counterpart, which is an electron within the electron cloud. And because they are counterparts, they annihilate. When they hit each other, they annihilate and that annihilation creates energy, as we all know, and it releases energy in the form of photons that for PET imaging, these photons are emitted in a 180 degree pathway from each other. And that's very important for imaging. One important concepts that I wanna highlight, just because it's gonna be important for us in our review of tracers, is that the length or the distance that the positron travels before finding the electron and annihilating is what we call positron range. And that positron range depends on the energy of the positron. The higher energy the positron has, the longer it's gonna travel before annihilation. And that has important implication for a special resolution as we're gonna talk in a second. And why is that? Is because what we're trying to image with the camera and the detector, we're trying to image actually the place where the positron originated, and that's in the nucleus. The far away you get from the nucleus to generate the photons, it creates a misregistration. So you're actually imaging something at this level, at the initial level, when you in fact wanted to do it or try to localize it to the nucleus. So the far away that the positron travels, the lower is gonna be the spatial resolution. We're gonna see some examples of that. The other important concept in PET tracers is that irrespective of the energy of the positron, the process of annihilation in the electron cloud is gonna generate these two photons. They are 511 KV, so it's a high energy photon, and it's more energetic. So you don't have a variety of energies that can create, can degrade the image quality, but it's actually a more energetic, high energy photons, which is gonna be great for imaging. Now, having said that, these are kind of the five elements in terms of physics and biology that we wanna cover and how that are important to understand when in the acquisition and the interpretation of images. So we're gonna talk about photon energy, we're gonna talk about half-life, we're gonna talk about positron range, we're gonna talk about tracer kinetics, in particular extraction fraction and retention fraction, and we're gonna talk a little bit about compartment modeling. So the first thing is we briefly touch upon that. So PET tracers, when they undergo the annihilation process, they generate, these two photons are high energy with 511 KV, which contrasts to SPECT tracers and that we see here in this graph that you have a 511 KV for the PET tracers compared to, let's say, technetium, which has an energy of 140 KV. So why is it important? It's because, as we all know, these photons are being generated in the body within the myocardium and they need to travel across tissue to actually hit the detector and be where it is finally recorded to form the images. So the photons have the possibility of hitting the myocardium in soft tissue and being attenuated and never reaching the detector. So that attenuation is affected by the energy of the photon. As we can see here in this graph, we have the percent of transmission in the y-axis and we have a column of water that increases as we go far to the right. And what we can see here is, and we have here two curves. One is, again, 511 for the PET and 140 for the technetium for SPECT. And we can see if at any single thickness of the column of water, what we can see here is that the percent of transmission, so the proportion of counts or photons that actually reach the detector is gonna be always higher with PET compared to SPECT tracers that are lower energy. And that comes very handy, especially in people that our patient have are high weight, high fat composition that can create troubles with attenuation, potentially creating defects that are just attenuation in nature. The other thing here just to briefly mention is that not so much in compared to technetium, but having a monoenergetic photon generation is great for creating ways to exclude scatter, which is gonna ultimately degrade image quality. For instance, when we use thallium, thallium has different peaks and that creates a more complex way of dealing with scatter, which at the end of the day degrades image quality. So the second concept, I think we again touch upon a little bit is the positron range. And we already discussed that it greatly influences the spatial resolution. And just as a way of reminder, we know that the larger the positron energy, the longer the positron is gonna travel before hitting the electron and annihilating to create the photons. And the longer it travels, the far away it's gonna be from the nucleus, which is what actually the place that we're trying to image and create misregistration, and which is gonna ultimately affect the spatial resolution, decreasing the quality of the image. So just as a matter of an example here, we can see in this picture that we have here rubidium on the left compared to flupidus. And we can see that the positron range for rubidium is in the order of 8.6 millimeters compared to flupidus, which is around a millimeter. And we can see here, let's start with flupidus. If we're imaging a rounder structure, as we see here, we see that with flupidus, we see a nice rounded compact and well-synchronized structure with flupidus. And compared to the rubidium, because the positron range is greater, and your localization of where the counts was generated is not as good. And we have this blurring and spillover of the counts creating this fussy description or representation of the nice rounded structure. So as you can see here, the spatial resolution is way better with flupidus, has a smaller positron range and a smaller energy of the positron compared to rubidium. And here's just in a depiction on now macroprofusion images. And we can see here, again, rubidium on the left compared to flupidus on the right. And we can see here is that flupidus first, you can have detect finer details, such as the papillary muscle. You can see that the blood pool is a little bit bigger because you have a better spatial resolution. In contrast, rubidium with a higher positron range and an inferior spatial resolution, you get more blurriness of the image and a smaller blood pool. It's not because the blood pool is smaller, it's because of that spillover and blurriness that relates to the longer positron range and the inferior spatial resolution within these two tracers. On the same token, the same spillover that you get in the blurriness can really obscure tiny proficient defects because of the inferior spatial resolution compared to a tracer has a higher spatial resolution and is able to detect more discrete proficient defects. Now, the third element that I think we all understand, but it's important to, again, go over is the half-life of the radio tracer. There are many half-life that we refer to, including the biological, the physical, the effective half-time, but the one that we commonly refer to in nuclear imaging is the physical half-life. And it's defined by the time that it requires for the radioactivity level of a particular source to go by half of that original value. And that's the time it takes. And here, this depiction on the graph, we can see that in this exponential curve, as you start with one activity, as you go over time, the amount of activity you're gonna get is gonna decrease by virtue of decay. And the time it takes for that activity to go by half, that's what is defined as a half-life. And why is that important? Well, it's important because most of the time that we do nuclear perfusion imaging, we don't necessarily image immediately after, sometimes we image a little bit later on. And how later you can image is a factor of the half-life. You have a very short half-life, you don't have much time to waste before getting the pictures because the tracer is gonna decay, and then you're not gonna have anything to image if you wait too much. On the contrary, you have a longer half-life, you have a longer imaging time window to get the pictures that you need. The other important concept is that if you have a tracer that has a longer half-life, it allows you to squeeze an exercise stress testing, inject the patient a big exercise, and then have the time to transport the patient from the treadmill into the camera and acquire the pictures. If you have a short half-life, and you don't have that luxury to inject the tracer, transport the patient and get the pictures. So as we're gonna see, some of the tracer we're gonna discuss today, given the longer half-life, allows for the performance of exercise stress testing. And that's an important characteristic to keep in mind. The second thing is, remember that we're not imaging one count, we're imaging the sum of counts over a period of time where we're actually getting the pictures. So you have a longer half-life, you have the flexibility to increase your imaging time. And by doing so, you can actually increase the amount of counts you're detecting, and that in general, improve image quality. And finally, the downside of half-life is that the longer half-life are associated with an increase in radiation. As we know, radiation exposure is gonna be a factor of the energy of the photon, but also as a factor of the time. So those tracers that have a shorter half-life are associated with a longer, sorry, a lower exposure to radiation, which is, I think is important. Now going into, you know, radiotracer kinetics, which is we're using more and more and is becoming essential in the evolution of patient with either at risk of with coronary artery disease. You know, what pharmacokinetics relates to is just basically the evolution of the movement of the radiotracer in the body. And we can not only see that radiotracer move in the body, but we can also measure it using time activity curves. And here on the panel on the left, what we can see is that we can easily assess how the tracer is moving through the heart. Tracer gets through the systemic veins into the right side of cavities. From the right side of cavities start crossing through the lung and to the capillaries into the left side. From the left side gets into the LV blood pool. And from there it's gonna be taken up by the myocardium. And that's at the point that we're gonna take the pictures of myocardial perfusion imaging. On the panel on the right, now we're not only visualizing how the tracers travel, but we're also counting, you know, the counts as they, you know, go through the different cavities across time. And we can see here is that if you put an area of interest in the RV, in the LV blood pool, in the myocardium, we can see is that, again, you know, the counts gets into the right ventricle and then they get pushed through the lung into the left ventricle. And as the left ventricle, blood pool of the left ventricle counts are going down as they are getting or taking up by the myocardium. And these are, here's important to make a little stop and highlight how that process work, because it's gonna be very important for us to understand some of the next concept we're gonna discuss. As we can see here, as the LV blood pool accounts are getting ejected into the order and then taken by myocardium, we can see that the myocardium start taking up the counts by a way of extraction. So in the first pass of the count through the myocardium, the myocardium extract the counts. And that's an essential process, that extraction. You want tracers that really extract everything that comes through or by them. But it's not only important to extract the tracer, you want a tracer that not only is extracted fully by the myocardium, but is also retained by the myocardium with minimal washout. And why is that important? Because when we actually take a picture of the myocardium, we don't take it necessarily early on when you're extracting, you actually wait a little bit to allow the blood pool of the LV to really have minimal counts. So you can have a big contrast between the blood pool and the myocardium. So you need, and you require the myocardium really retains the tracer to actually get a nice quality picture of the myocardial perfusion of the patient that you're evaluating. So I think these are two concepts that are important to hold on to, which is extraction and retention of the tracer when we talk about kinetics. So now going a little bit deeper. So extraction fraction. So when we talk about extraction fraction, we're basically referring to the proportion of the tracer that is taken by the myocardium in first pass. And we talk already that this is essential. What we can see here in this graph is not all the tracers have a perfect extraction. We can see it here with uptake rate in the Y axis, which is a factor of the myocardial blood flow and the extraction fraction. And on the X axis, we see a myocardial blood flow. We can see here is we want a tracer that has a linear relationship between the myocardial blood flow and the extraction fraction. So that way, if your myocardial blood flow is one, your uptake rate is gonna be one. So you have a linear relationship. So it's easy to back calculate myocardial blood flow. But unfortunately, all the tracers are not like that. We can see here, for instance, that rubidium has a linear relationship early on, a very low flow. But as your myocardial blood flow increases, it reaches kind of a sort of a plateau. And now your myocardial blood flow doesn't equal your uptake rate. And that's a problem for, and when you're trying to calculate myocardial blood flow, because now your uptake no longer equals your blood flow, and you need to do corrections. You need to do mathematical correction to make this line from this kind of plateau line to now aligned relationship. And that can include errors. So that's why extraction fraction is essential, especially when we're dealing with myocardial blood flow, because the more linear relationship you get, the easier it's gonna be to back calculate the myocardial blood flow from the uptake in the myocardium. As we can see here, O1R15, which is the gold standard for myocardial blood flow quantitation, has a linear relationship followed by fluorpyridase and plus by ammonia. But as we already discussed, rubidium is inferior in that particular aspect. This plateau of the curve, as the flow or myocardial blood flow goes higher, is what is called Roloff phenomenon. And here, just to kind of emphasize again, you know, if your extraction fraction is 100%, meaning every single count that gets through the myocardium in first pass is taken up by the myocardium, well, your uptake in the myocardium is gonna be equal to myocardial blood flow. So what's the relevance of the extraction fraction? It not only affects or is important for quantitation, but it's also important for the quality of the picture. And why is that? Well, you know, we see here that, again, the quantitation is way more accurate when you have a linear relationship, you have a 100% extraction fraction because you avoid mathematical corrections. But you also have an improvement in the identification of defects. And why is that? Well, let's take out here a moment to go back to the graph. And what we can see here is we have water 15 on the left and we have rubidium on the right. And we know that water 15 has a great performance in terms of extraction, it's 100%. So what we have here is a patient that, what we did is we create, we ligate the LAD and create a narrowing in the LAD, compromising the capacity of the blood flow to really augment there. What the rest of the myocardium is normal. So if the normal myocardium in this particular example is, the normal myocardial blood flow is around 2.4. And by way of creating that narrowing of the LAD, we reduce the stress flow in the LAD by 25% down to 1.8 miles per meter per gram of tissue. So if you use water and you plot using this graph, you say 2.4, it goes to 2.4 in uptake. And then from 1.8 of flow gets to 1.8 in tracer because extraction fracture is 100%. You're gonna see that that 25% reduction in flow translating to a 25% reduction in the uptake. So you're gonna have a 25% contrast difference between those tissue when you look at it. In contrast, if you talk about rubidium, for instance, and you get exactly the same example, you have the same patient with LAD narrowing, you have a reduction in the flow during stress of 25% along the LAD territory. And your image now with rubidium, what you're gonna see is you're gonna get that plateau and roll off phenomenon with rubidium because the extraction fraction is not 100%. And when you plot how much is gonna be the difference in tracer uptake, you're gonna see that that 25% difference in flow is gonna translate now into only a 10% contrast difference in terms of tracer uptake. So it really decreases your capacity to identify a perfusion defect. And the higher the flow, the microbial flow, the lower is gonna be your capacity to really discern and identify this perfusion defect. So, and that's an important concept to have in mind in how having a high extraction fraction really improve the sensitivity, but also the true severity of a perfusion defect. The last two important points about extraction fraction is that the more you extract into the myocardium, well, the more count you're gonna get in the myocardium per dose injected. So the more counts you get in the myocardium with less counts in the blood pool and in the background, the better it's gonna be your image quality. And because all the counts are being taken by the myocardium, you know, hopefully in the first pass, other organs like the liver might not have the chance to take too much. So your myocardium to background ratio really improve. And those are things that really makes a big impact on the quality of the image. Now, moving on to retention fraction, which is, again, the second concept. We already discussed that you need to extract, but also you need to retain in particular for when you're trying to get myocardial perfusion images. So the retention fraction is the proportion of the trace that has been extracted on first pass that is now retained by the myocardium over a predefined period of time. And that's a factor of the washout. So you want the traces to go in, but not to necessarily wash out and get out. And as we discuss the accuracy of the myocardial perfusion imaging, and we're talking when I say myocardial perfusion imaging, I'm referring to the static images, not to the quantification of flow. And we have here a representation of different tracers. None of them have ideal retention. You know, perhaps repeated pass here is the best that has the best retention fraction of all the PET tracers and even some of the SPECT tracers. And we see the rubidium is also inferior compared to repeated pass and ammonia. An important point here is that, although water was beautiful for extraction and so is perfect for quantitation. So water, because it freely diffuses, we see that the retention for water is zero. So the tracer goes in and goes out immediately and wash out and equilibrates. So that's why the, although it's perfect for quantitation of flow, for a static images and for perfusion images is the worst, because you don't retain anything. So by the time you image, everything is already washed out. So what's the relevance? We touched upon a little bit. So your quality and accuracy of your perfusion imaging, the static perfusion image is way better as long as you have a good extraction, but also a good retention. And this is an example here. We can see on the left, we have a tracer has perfect extraction, but also have a great or close to perfect retention. And remember that we extract, we don't image right there. We wait a little longer just to make sure that we have no counts in the LV cavity and we take the pictures. If we take the pictures here, we're gonna see there is a lot of counts in the malachorium, but it's nothing in the blood pool. So you have a nice contrast between the blood pool and the malachorium. We get nice quality images because we have a high retention fraction. In the counterpart, we see here a low retention tracer. In this case, it's water. And as you see here, it has a perfect extraction, but right away after extracting everything, it's gonna re-equilibrate with the blood pool and it's gonna wash out quickly. So even if you try to image this patient right away, quickly after you inject the tracer, first, some of that tracer is already gonna go down because it's already washing out by the time you take the picture. So your quality gets lower, less count in the malachorium, you get more hotspots, which is not good for image interpretation, but also your LV blood pool still have some counts. So the contrast between the blood pool and the malachorium is not great. So it results in an inferior and non-diagnostic picture. So that's why it's important to have a high retention fraction. Again, similar to extraction fraction, the more counts you retain, the more count you have to count, the better the quality of the pictures, and then the less you're gonna have counts available for other organs to take, similar to what we discussed with extraction fraction. And this is just kind of a depiction on how it's important to understand these concepts because what we're trying to get at as a perfusion tracer is try to create a pictorial representation of the flow. And in this depiction, we have flow, which shows that it has a defect along the anterior wall. And because these particular tracers that we use doesn't have a perfect extraction fraction, we can see that perhaps the size, but also the intensity of that perfusion defect goes down because it doesn't have a perfect extraction fraction. And then because there's some washout, when you actually finally get your Markov perfusion imaging after a certain period of time, then you are underestimating the size, but also the severity of the perfusion defect. And in occasion, if it's a very subtle defect, you can miss it altogether. So that's why tracers that have a high extraction and a high retention really have a higher sensitivity for the identification of perfusion defects, especially those small and subtle defects. And this is particularly true at higher flows because the higher the flows, the higher it's gonna be your washout and the inferior is gonna be your extraction fraction. So you're more likely to miss defects at higher Markov flow than lower as we saw. Now, let's talk briefly here about compartment modeling and without intention to really get into the physics and all the mathematical kind of calculations. But I think the first thing to understand is that when we talk about compartment modeling, we're just basically describing or quantifying a process that might vary in time, but not in space. For instance, one particular compound gets transformed into another one, or a particular compound is being retained by the mock arm. So the quantification, not only the visualization, but now the quantitation of that process, which is very relevant, that's what we do with modeling. So when we refer to compartments, what we're referring to is that, there are the different unique states that are radio tears can be within the same space. So the traces can be in the blood pool, or it could be within the myocardium, the traces can be freely free in the cytoplasm, or it could be bound to a particular compound in the myocardium. So we can really quantify those states with the use of compartment modeling and the classic example of how we use these is for myocardial perfusion quantitation. So basically this is a simple schematics of a model. Basically the model has, you can describe it as a matter of how many compartments you have. So you can have one tissue or two tissue compartments, depending on how the tracers get transformed or get bound to things and so forth. So we're gonna use the simplest of them, which is the single compartment model. And the way it works is basically, when you check the count, the count goes in into your blood pool and that blood pool is the input function. That's what brings the count. And from that input, the myocardium can either take up tracers and get it into the cell or into the interstitium within the myocardium. And that tracer can also get back and return to the blood pool as a matter of a washout. If you allowed enough time, what's gonna happen is that the system is gonna get to an equilibrium and you're gonna have the same rate of uptake. It's gonna be the same rate of washout. And we said that the model is now in equilibrium. And when you reach that state, you can calculate the volume of distribution, which is basically the ratio between the concentration in the tissue divided by the concentration of the blood. So by using or understanding and knowing this volume of distribution and knowing the concentrations of the tracers over time in the blood pool and the concentrations in the tissue over time, you can use those factors to actually derive what's the myocardial uptake is. And once you have your myocardial uptake, again, if you have a 100% extraction fraction, you can say that your uptake is equal to your myocardial blood flow. If your extraction fraction is not 100%, then you do corrections to make that relation linear. And then you can derive your myocardial blood flow as a fact of milliliters per minute per gram of tissue. And that's how a compartment model works. So an important point to make here is that then your calculation of your input function is crucial because any errors in this calculation creates errors in your model, of course. Here in the second, just a little brief mention. We can see other tracers behave as a two-tissue compartment like we see with ammonia. And it's basically, the tracer can get into the cell or to the interstitial, but also can be trapped into the myocardium by incorporating two different molecules. And you need to account for all these different changes in the tracer to get an accurate quantitation. So again, this is now used very often in clinical practice. And we can use compartment model to actually derive and back calculate what the myocardial blood flow is. And that becomes, that's a very powerful tool for not only diagnosis, but also for prognosis of patient with suspicious for or a known CAD. And there's compartment model for like metabolism, but also for amyloid plaque quantitation and I'm pretty sure this will be covered in greater depth in the flow quantitation module. So the last important concept to cement is that as we talk, what we're trying to get is to try to determine the quantity of the myocardial blood flow. In this particular case, when I go from left to right, we see that there's a proficient effect again in the anterior wall. Assuming that your extraction fraction is perfect. So you have a very nice representation during uptake. But then when you try to image the patient for macro proficient imaging and your retention is not ideal, you see that the severity and perhaps the extent of the proficient goes down a little bit. So you underestimate a little bit of the severity of the proficient defect by way of looking at the images. And this is the image that you get. Again, if you compare that to flow, you are underestimated again, severity and perhaps extension. So by using myocardial blood flow quantitation by way of using comparative modeling, what you can do is you can recover the severity of the proficient defect. And you can now analyze not only the images, relative images of the proficient, but also you can see how the flows are behaving and you can really recover the true proficient defect contrast. So what are the changes of kinetic modeling? And for quantitation, I think it's important to have that in mind. I mean, an accurate arterial input function is essential. And here there are two examples. And the first one is that if you get truncation of the peak, so in order to get a nice arterial input function, you wanna have a pretty compact and peak left ventricular blood pool curve to really get an accurate quantitation. So if you have truncation, because perhaps sometimes you give too much activity and that really saturates the detectors, the detector cannot count more. So when you do that, that you might overestimate a little bit of the flow. The other thing you need to be careful is that you really need to have kind of a normal arterial input function curve, which is again in green here. We can see we have a double hump and we have blood flows that are higher than the counts in the LV blood pool, which is not possible. And what we have here is that, this double hump is just a reflection of basically the presence of a shunt. We can see here, we can see that from the systemic veins, we're gonna see when it goes around again, we're gonna see that it goes here, systemic gets to the LV and then gets recirculation. So this is a case of a fistula between a systemic vein to a pulmonary vein that creates this double hump. So essentially you cannot interpret or use these for quantitation. So where are the strengths? Well, you can quantitate things and you can recover proficient defect severity because you're going back to the quantitation of flow. But on the other side, you are very sensitive to errors because if your mathematical corrections for tracer doesn't have a perfect fraction, can create errors. If you don't measure your arterial input function properly, as we saw in the last two cases, then you're gonna have a number that is not true in terms of quantitation of flow. So now that we understand a little bit some of the basic concept of tracers. So what's an ideal pet proficient tracer? Well, it's one that is available as a unit dose from a regional secretron. So we all know that some of these tracers are created in a cyclotron. And so that has been a very important issue in terms of getting access to pet. It was not only about having the actual camera, but also having the access to the tracer. And for instance, like for ammonia, you need to have the cyclotron in house. So if you have a tracer that you can have as a unit dose that someone delivers to you and that you can use, that would be really helpful. And that's kind of a characteristic of an ideal tracer. You wanna have a low position range because that improves your special resolution. You wanna have a high extraction fraction and a high retention fraction for the reason that we already discussed. You wanna have minimal retention within other organs to avoid either overlap or skewing defects and have a high Markov-to-background ratio. You wanna have a longer or a relatively longer half-life so you can squeeze some rest exercise imaging and not just a pharmacological stress testing. And you need to have the appropriate characteristics as we discussed to have the most accurate quantitation of flow. And it's gonna be a compromise between how long your half-life is gonna be to allowed exercise, but at the same time, not too long so you don't get too much of radiation exposure. And as with anything, this ideal perfusion tracer doesn't exist, at least now. Hopefully in the future, we can come up with something. So now when we talk about perfusion tracers, these are the fours that you can read about. And only rubidium and ammonia are FDA approved for clinical applications. Water 15 is used for research purposes and for PDDAS is being studied in phase three trials with promising to be a great tracer. So now we're gonna get diving into a little bit of the characteristic of each of the tracers. As we already discussed, rubidium has been approved for clinical use. And basically rubidium is a potassium analog. And basically it gets taken up by the cell via the sodium potassium ATPase, like the same way that the potassium is and the same way that the tallium works. It has a short half-life of around 78 seconds. So that results in a low radiation, but it doesn't allow you to do exercise during the stress imaging. So the way that rubidium is delivered is that the parent radionuclide is generated in a cyclotron is put into a generator. And that generator is given to the site. And basically in the generator, what happened is the strong sodium 82 decays into rubidium 82, and then you can elude or milk the generator and give the dose that you need. So this generator has a shelf life of around four to eight weeks. And this is a good way to increase access to PET because you don't need to have a cyclotron in-house. You only need to kind of buy the generator. The other thing is that, you know, you can't milk or elude the generator quickly every 10 minutes. And that really promotes throughput. You can get cases after cases because you don't need to produce something and that's all you have for the day. So you can, you always have rubidium that you can get out of the generator every 10 minutes. So you can, you know, create atoms and, you know, lay patients and you always have the tracers to be ready to be used. So as I said, no need for onsite cyclotron. And the way the generator works is basically you have this recipient, which is shielding the core. And in the core, you have a column and that column has tin-tin oxide. And that tin-tin oxide binds to the strontium 82 with high affinity. And as a strontium decays into rubidium, rubidium doesn't have that much of affinity for the column. So if you flush with saline, you're gonna take out and flush the rubidium, but not the strontium. So what comes out of the outlet when you elute, the generator is only rubidium. And of course you need to check regularly on a daily basis, make sure that that strontium is not being eroded or flushed along with the rubidium because that could create troubles in terms of your image quality. But that's basically how a generator works. You have it attached. You can have it attached behind the PET and study has demonstrated that using this kind of more automatic elution system in which it's just a computerized system with saline and you basically click and push how much activity you wanna inject to the patient. And it works in an automatic way, just attached behind the camera and the patient get injected and you get your pictures. And this has been found to be a very reproducible way of injecting very accurate amount of activity. And so this model, as you can see, so you can do it every 10 minutes, you can do it in an automatic way, very accurate. So it's really favor in centers that have a high throughput and high volume of patients referred for PET. So this is kind of the classic example of a PET protocol. As we said, the short half-life of the rubidium only allows for pharmacological imaging and not stress imaging. And you can really image very fast stress after rest because of the short half-life. And your radiation is around one to two millisieverts. So rubidium, when we talk about quality of myocardial perfusion imaging, given that it has a low extraction fraction and a low retention fraction, as we see here, the contrast resolution is not as great as other tracers. Has a longer positron range, so the special solution is not great either, in comparison to other tracers. And such as a lower background to macular to background ratio with uptake in the stomach and the liver and so forth. But despite all those limitations in comparison to other tracers, Rubidium have demonstrated to have a high accuracy for the detection of obstructive CAD with excellent prognostic value, despite the limitation that we discussed. And we have here a depiction of a case with a proficient effect along the LED, perhaps representing diagonal territory, showing a reversible proficient kind of defect. You can see how nice you can detect, despite the blurriness and the inferior resolution. So when we talk about quantitational flow, again, not an ideal tracer due to lower extraction fractions. So as we can see, we need to, after doing all the modeling, we need to correct that curve to make it linear again, so we can actually derive flow. One or two tissue compartments can be used for Rubidium. Given that they have a short half-life, so you need to really give a high dose, so you can have enough dose to actually image the patient once the blood pool counts has decreased. And because of that really increased amount of counts that you sent, that can saturate the receptor, and that can create that truncation of the internal input function that we already described, creating troubles in the interpretation. But again, so despite all these kind of inferior characteristics, quantitational flow has been reproducible with Rubidium, although it has a little less accuracy and more variability compared to better tracers such as ammonia and water, in terms of perfusion or microblood flow quantitation. Now switching to ammonia. Ammonia is produced in a cyclotron, so you need to have a cyclotron in-house. So that really creates limitations, because you need to have the personnel, you have the equipment, have the cyclotron working to really generate. And you need to really plan your day ahead and say, I'm going to do 5, 6, 7, 10, 15 PET studies, because you need to really time and secure the production of ammonia for the day. And that creates trouble, because if all of a sudden you have an add-on at the end of the day, and you didn't account for that, then you reproduce the ammonia for the day, so you cannot really add that on. And on the other hand, if you have a patient that was scheduled for a particular time, and your tracer was delivered, and the patient, for whatever reason, got delayed, and as we see here, the ammonia has a half-life of around 10 minutes. If the patient arrives 10 minutes late, an hour late, then your dose is gone. So you cannot image the patient, so that spot has been missed, and you cannot create more ammonia. So it really creates challenges in terms of operation when we use ammonia. With now the proposal of a mini cyclotron, in which now instead of having this big particle accelerator, we can have a smaller equipment with less shielding in a compact and semi-automatic way, there's a promise that some of these resources can be generated a little faster, and can improve access to areas in which a standard cyclotron is not feasible from a financial or from a space and from the management perspective. And some of the study has been very promising, suggesting that the quality of the tracer production is good, although more studies are needed. But stay tuned in terms of the production of these mini cyclotrons that can improve access. So ammonia, the way it works is basically diffuses across membranes in a passive way. And as you can see here in the graph, it can go in or it can go out, and finally can be a trap by incorporation of the ammonia into the amino acid blow pool. So you can already infer that ammonia behaves more as a two-tissue compartment by not only freeing the cell, but also being incorporated and trapped into the amino acid blow pool. And finally, those are metabolized via the uracycle and is eliminated through the urine. So this is a typical ammonia protocol. One important and unique characteristic of ammonia is that given to have a longer half-life of 10 minutes, then you can really squeeze an exercise. You can really do a ramp protocol and get the patient to peak heart rate quickly, and then inject the tracer in the treadmill or in the bike, and then get the patient quickly into the camera and start acquiring your stress images. And so there's a possibility for that, and that's what we use when you're looking for, let's say, the significance of anomalous origin of the coronaries, when you really need that exercise and pharmacological is not enough. An important concept here that I think I'm pretty sure that has been touched upon in other modules is that although we can do exercise, still the quantitation of flow when you do exercise is not possible, given that you need to really quantify that counts going through the RV to the LD. And the only way you can do that is by having the patient in the camera. So there's ways to try to kind of go around that, but in a simplistic way, ammonia allows for exercise, stress testing, and with that, you can actually look at the images of macrophage and imaging, but not actually quantify flow. You can only quantify flow if you do a pharmacological one. One of the challenges with ammonia is that because you have a longer half-life, sometimes you need to either wait longer between rest and stress, which you create a longer imaging time and decreases throughput, or you can do what we do in SPECT and just say, well, I'm gonna do three times a dose in distress to kind of resolve the shine-through phenomenon and not have the longer acquisition time. So you can decide if you wanna wait or just kind of go three times a dose for the stress at the cost of a little higher radiation. And here we can see the radiation is in the order of 2.4 to 6.1 millisieverts, given the longer half-life. Now, when we talk ammonia about, you know, ammonia and macrophage and imaging, it's a better tracer, better extraction of around 80%, retention fraction of the 60, 70% range, and that improves, of course, contrast resolution. You're able to see better perfusion effect that it resemble or kind of correlates better with the perfusion. So it has a lower positron range, so higher spectral resolution. The positron is around 2.5 to 6.1 millisieverts. And because of the high half-life and the high retention fraction, you have higher count statistics. You have more counts to be counted that improve, again, image quality with less uptake by the background. And again, similar to rubidium, it has a very high diagnostic accuracy and outstanding prognostic value. And we can see here image, macroperfusion images of ammonia. We can see there is very nice and discrete borders with higher resolution compared to rubidium. We can see a little bit of spleen, liver uptake in the picture. One of the challenges with ammonia, and this has been talked a lot, is just the lateral wall defect, which is, again, the fissioli, or the reason for that is still kind of unclear. But we can see here on the top in hot metal scale, we can see that the lateral wall is a little less hot compared to the septum, and that's usually not normal. You always, your lateral wall tend to be a little hotter than your septum. And you can see that the same in the same patient on the gray scale, which is more a linear scale. And we can see here, there's, you can detect there's a little bit of a ghosting. And that ghosting is what, you know, some people believe is that creates that defumination of the count, creating this illusion that there is a profusion defect in the lateral wall. And that's a common challenge that, you know, ammonia presents when we're reading, but it's very easy to identify by ways of detecting this ghosting, and by way of looking at quantitation of flow, and also looking at the gated pictures. The other thing that ammonia challenge clinicians and readers is that it can be taken up by the lung. On the left side, we can see that, you know, there's lung uptake at rest that gets kind of better when you look at the pictures doing stress. So that doesn't represent that there is ischemia or anything, but this is something that is often seen with ammonia in patients that are smoker and have lung disease. And what we're seeing here is not that there is less with stress, we're seeing here is that the myocardium has more. So in relation to the myocardium, this look relatively decreased, but the uptake or the absolute uptake in the lung is the same. And that's something that we see with ammonia and that can create troubles, not only with smokers, but also when a patient is in heart failure and there's water in the lungs, ammonia goes directly into the water and that can create challenges for interpretation of the pictures because you have some lung uptake, you see in the right top corner, there's some lung uptake that overlaps with the lateral wall and that can obscure upper fission defect. The same for flow quantitation, some of these flow that is going into the lung is being counted into the lateral wall, falsely inflating the flow in the lateral wall. So that can create a challenge in interpreting those segments that counts as being spilled over from the lung. For quantitation, ammonia is excellent, highest fraction fraction with minimal roll-off phenomenon. It behave as a two-tissue compartment and has been extensively validated with microsphere, which is a gold standard across a wide variety of flow ranges and has a high accuracy with less variability in quantitation compared to rubidium. So here, just kind of a quick depiction of ammonia versus rubidium. And we can see here, we start with ammonia and we have the flow. And as we go into the extraction, it's a very good extraction, but then a little decrease in the perfusion contrast, the perfusion defect contrast because of the retention. So the step down is at the retention compared to rubidium in which you have stepped down in the uptake because the extraction fraction is slower, but also during the retention, making ammonia a little better tracer from this particular perspective. So water, we talk about it's not an FDA approved. It's producing an in-house cyclotron and perhaps the use of these mini cyclotrons can be of potential utility. Has a very short half-life of 2.4 and has an intermediate special resolution given intermediate positron range. And what we all know is water diffuses freely and rapidly between membranes and is metabolically inert. So it can go in and can go out quickly. This is a classic image protocol for water 15 and very quick image acquisition because of the short half-life and has a low radiation again because of the short half-life. So this is the duality of water is that it's the best for quantitation because of a perfect extraction fraction. It only has one tissue compartment mold making it simple and you have a very good, as a consequence, a very good correlation with microsphere in terms of quantitation of flow. However, it's the worst for more color proficient imaging for static images. And why is that? It's because the retention fraction is zero. So anything that comes in quickly comes out and equilibrates. So you don't have any good counts in the malcartium and there's a lot of counts in the blood pool because of the washout and the equilibration of the counts. So that's why water, although it's great for quantitation is not useful for getting actual images that you can interpret as we do for rubidium and ammonia. Unless you do fancy correction for backblower activity or some other corrections. And again, this is the example that we already described before. The washout really compromised the capacity for us to actually get accurate images to look at regional profusion. So you have to rely only on quantitation and that's why it's used mainly for research. So what people has come up with is that saying, well, why don't we create a color coded representation of the flows instead of imaging later on and get the pictures or static pictures. Why don't we just convert the flows into a color coded or color scale kind of depiction and that's what people have done. And that's what is called parametric imaging with water 15. And we can see here's a normal, we have a stress, sorry, a rest acquisition here and then augments with stress. You can see how the flow of bends in the different segments and this is a normal. In this other case, you see that it's a normal rest profusion. Again, you see that the colors changes but it's because the scale is what is important. So this reflects or translating to a flow in kind of 1.8 range, which is within normal limits. However, when you augment the LAD territory, you see that there's no augmentation or no significant mutations in the one range showing a profusion defect along the LAD. And finally, multi-vessels have a relatively or a normal or actually normal to high flow at rest. And then you have no augmentation whatsoever in the range of one during stress. So people have been using parametric imaging with water 15 to get around the issue of having low retention. So finally, a few words on fluorpyridase, which is kind of the new kid in the block, which has been promised to bring great things to the field. It's currently in phase three clinical trials, of course, not FDA approved. Fluorpyridase is an analog of the insecticide pyridine. And what it does is basically from the blood pool gets into the cell and then binds to the mitochondrial complex one, which is in competition with the ubiquinone and there it becomes trapped. And it has minimal or washout. So as a constant has minimal or has a, sorry, a good retention fraction. And it's clear through the liver, GI and kidneys. So it has a longer half-life of 109 minutes. And that's great because now we can just do unit dose. We can actually create the tracer in a regional cyclotron and deliver the dose as a unit dose for areas that doesn't have the capacity to have a cyclotron. It has a great positron range, has the best positron range of one millimeter or so. It has excellent special resolution, has a high structure and fraction of 94% with a long half-life. So high count statistics and high contrast resolution and with a high myocardium to background ratio and has negligible roll-off phenomenon. And we can see here is very close to what water 15 has in terms of structure and fraction. So very promising characteristics. This is a depiction of how flopidias can look, very nice delineation between the blood pool and the myocardium. And you can identify some of the papillary muscles. So great image quality. And in this particular case, we see that there's a reversal perfusion defect involving the mid to kind of distal kind of inferior wall. So this is the protocol has been used in clinical, sorry, in research protocols because of the long half-life, you either have to wait a long time before you have the decay of the rest to give the rest, or you can use the one to three dose ratio protocol to combat and get around that, the shine through phenomenon from rest to stress. And you can minimize imaging time, but you might increase radiation. So the radiation is in the order of 4.8 to 6.4. And that opens the capability of doing two day stress and maybe just doing stress first, as we do with SPECT and avoid the rest if the stress is normal. So the data from phase three clinical trials so far have shown that some mixed results, although the sensitivity was higher for the detection of perfusion defect compared to SPECT. And we expected that given the better characteristics of this tracer against the SPECT tracers, but it did not meet the specified non-inferiority criteria in terms of specificity. And that might have to do with how sensitive this thing for subtle changes in blood flow that might not correlate with obstructive epicardial or might be related to more subtle changes in flow. So that has to be taken into account. So have a superior discrimination of obstructive CAD compared to SPECT and based on this phase three trial. And we need more data and it's currently ongoing a second phase three trial sponsored by GE. So this is just a slide summary that we're not gonna go. We have already cover each of these elements. A second table here, just to show some of the validated cutoff for normal perfusion and for normal stress flows and Magal flow reserve for ammonia and water and rubidium. We don't have anything for PDS. And finally, just to summarize in terms of strengths and weaknesses, if we are trying to get an exercise, Magal perfusion imaging with PET, ammonia is just a current tracer or standard tracer for these because of the longer half-life. So Flopida with a longer half-life is promising. We'll see if that pan out to be a good tracer as we all want that to be. In terms of patient throughput, if we're gonna get a lot of volume, be able to accommodate for atoms and so forth and rubidium is the winner compared to ammonia and Flopida is that you need to plan ahead better and it's more like a more challenging operations and doesn't allow for that high throughput unless you have planned for it. In terms of quality of the pictures of the static and perfusion pictures, in theory, ammonia and Flopida has a better quality given the better extraction fractions and better retention fractions and compared to, let's say water, which has very poor quality because of the poor retention or the zero retention that it has unless you really adjust for those things with the parametric images we discussed. For quantitation of flow, water is the winner. We'll see what Flopida shows followed by ammonia. In terms of weaknesses of water, then we already discussed that, the poor retention fraction really makes this tracer not useful for static images. Rubidium has an inferior resolution and significant roll-off phenomenon with the implication that it has that we already discussed. Ammonia, it's important to understand the lateral wall defect and the lung uptake and the challenge with operations. And Flopida, we'll see what that turns out to be in terms of clinical performance in clinical practice. Radiation, of course, the longer the half-life, the more radiation you're gonna get. So ammonia and Flopida has incurring in more radiation to the patient. So you need to really be very careful with your protocols to try to minimize that. And as we already discussed, only rubidium and ammonia has been approved. So just to finalize, there are here some take-home points that is important to emphasize. And some of this is gonna be repetition. I think repetition is the best indicator that there is. So the higher accuracy of PET imaging overall over spec in part is explained by the characteristics of the PET tracers. Not only all about the technology and the camera, but also the tracers with high energy, more energetic, better performance and the unique characteristics that we discussed really give the advantage to PET. So higher energy tracers have less attenuation. The shorter the positron range, the better is the spatial resolution. Ideal half-life of PET tracers is a compromise between the capability of doing exercise and having access without having a cyclotron in-house, but also trying to not have too much of radiation and be able to get people through the image quickly without too much of a waiting. An ideal PET tracer should exhibit a high, if not perfect fraction fraction and a high retention fraction with negligible role of phenomenon that improve quality of the picture, but also quantitation of flows. And I think it's important to understand that although some centers use one tracer or the other one, I think it's important that we all understand the strengths and the weaknesses. And I think as we move forward to the future, I think it's important to recognize them because these needs to be used in a patient center strategy in which we use one tracer or the other one depending on the needs and the diagnostic purpose of the test. For instance, if you're gonna do an exercise for an anomalous coronary, you need ammonia. If you're looking for throughput and everything and are short in image equality, lower radiation, maybe you use something like kind of rubidium. So important to kind of adjust the tracer also to the patient in front of us. So with that, I wanna conclude and wanna thank so the organizers of this amazing PET curriculum for the invitation. And it was an honor to participate in this.

cardiovascular PET tracers

rubidium

ammonia

water

fluoropyridaz

FDA approved

clinical use

research purposes

phase three clinical trials

exercise stress testing