So let's get going. Thanks for the introduction, Daniel. My objectives today, at the end of this presentation, you should be able to explain extracorporeal membrane oxygenation, recall protocols to improve CT imaging of the pulmonary arteries, heart, and aorta in patients on ECMO, and anticipate and recognize imaging artifacts for accurate diagnosis. We'll start with a bit of background. CT imaging of the cardiovascular structures in patients with severe heart failure and or arrhythmias or respiratory failure requires thoughtful protocol adaptation to obtain diagnostic studies and avoid imaging pitfalls. And ECMO is a modified cardiopulmonary bypass to support life, permitting further treatment and recovery during severe cardiac or pulmonary failure. And there are two types. The first is venovenous ECMO, and the second is venoarterial. Now, with venovenous ECMO, this is used in patients with respiratory failure. Their cardiocirculatory function is normal or near normal, and blood is taken from and returned to the central venous system. And the aortic bloodstream flows in a normal direction. So important to remember that it flows in a normal direction. And here is an example of a patient with severe COVID pneumonia on VV ECMO. The chest x-ray on the right shows the venous cannula projected over the RA-IVC junction. The blood flows into the pump and then the oxygenator, and then it is returned to the systemic venous circulation through a large cannula in the SVC. Venoarterial ECMO, on the other hand, is used in patients with acute refractory cardiac or cardiopulmonary failure, and blood is withdrawn from the venous circulation via a central vein or the right atrium by a gravity siphon. And the blood is returned to the arterial side of the circulation, typically to the aorta. And there are two types, central and peripheral. Central VA ECMO is established in the operating room after cardiotomy when the patient is unable to be weaned off cardiopulmonary bypass. The inflow and outflow cannulas are placed directly in the right atrium and the ascending aorta, respectively. And the direction of the reintroduced blood is anti-grade, like a physiologic circuit. So that's important to remember. And here's an example of a patient on a central VA ECMO. You can see the drainage catheter is projected over the right atrium here. The blood flows outside into the pump, into the oxygenator, and then is returned via a large cannula that's placed in the ascending aorta. So the blood that's returning is going anti-grade in the aorta. Now, with peripheral VA ECMO, the situation is different. Femorofemoral configuration is the most common vascular axis in adults. Here, a large bore inflow cannula withdraws blood from the systemic veins, and the outflow cannula returns oxygenated blood to the aorta in a retrograde direction. Important to remember that. And this is established outside of the operating room. And here's an example, a scout view from a CT. We can see the large venous cannula, outflow cannula, with the tip projected over the right atrium here. It's exiting via the left femoral vein. Blood is entering the pump, and then the oxygenator, and is returned into the arterial circulation through this large femoral arterial cannula into the abdominal aorta. And the blood returning is going up the aorta in a retrograde fashion. So there is lots of imaging challenges in patients on ECMO. Just a show of hands here, has anyone done any imaging of the chest or abdomen in patients on ECMO? A few of you have, great. All right. Well, for the rest of you, I hope you'll learn a lot about how to image these patients appropriately. So the first challenge that you have is that the contrast injected into the venous line might be siphoned off from the anatomy of interest by the ECMO circuit. So the result is contrast dilution and non-diagnostic studies. The altered hemodynamics and the flow for patients on peripheral VA ECMO really kind of screws up your timing for your contrast bolus. It really makes it difficult. And then you have to deal with the CT table excursion, which can be a significant risk of dislodgement of the ECMO cannulas or other life-saving catheters. So when you are asked to do a study on one of these patients, there are six steps that you need to follow. And the first one is really to ask the clinical question, what will doing the CT study answer the clinical question? If it's yes, will it change patient management? And if both of those are a yes answer, then you really have to work with the ordering provider to plan for transportation and acquiring the CT images. The second step, once the patient has arrived to CT, you have to consider the excursion and the speed of the CT table to ensure that there is an adequate slack on all the life-supporting lines. All metallic objects, of course, need to be removed from the region of the examination. And then key here is to do a slow test of the table excursion prior to the scan, just to make sure you have enough slack on all those life-saving catheters. Then you have to think about contrast injection considerations. So where are you going to inject the contrast from? The easiest, of course, is through a peripheral IV in the antecubital fossa. But you have other options noted here. Note that you can inject into the oxygenator in the arterial or venous ECMO cannula. You have to think about the duration of contrast. You have to think about the time that's needed for the contrast to opacify the piece of anatomy that you want to see. So the injection site, the native cardiac function you have to think about. Is it good? Is it really bad? And the flow velocity of the ECMO circuit. And then one thing you should think about is increasing your concentration of iodine in your contrast, because a lot of the time you're going to be really counting on recirculation of that contrast and doing a second delayed acquisition. The other thing is you want to inject a lot. Increase your volume to three mils per kilogram. Another thing you need to think about is the ECMO. Does it have to stay on, or can it be turned off? This you need to discuss with your clinical provider and the perfusionist, because there are risks of clinical deterioration, oxygen desaturation, and clot formation from stasis. So if you can pause, how long do you need to pause? And that's going to be based on the rate and the volume of contrast you're injecting, the catheter length, and the time needed to image that anatomy. If the circuit can be paused or slowed, great. You can inject the contrast through a standard catheter. If you can't pause it, the contrast will have to be injected into the oxygenator or the arterial inflow side of the ECMO circuit in conjunction with the perfusionist, because then you want to avoid air bubbles in the tubing. So now let's start talking about the anatomy that you really need to examine. So if you're looking for clot in the pulmonary arteries, for example, the challenge is that the blood, and thus the contrast, might be siphoned off by the ECMO circuit and bypass the pulmonary arteries completely. So the solutions best is to temporarily pause or reduce the ECMO flow to allow the contrast to go past the outflow or venous cannula and opacify the pulmonary arteries. You can also inject into the oxygenator with full ECMO flow for patients on VV ECMO, or do a retrograde direct injection into the venous cannula for patients on VA ECMO. You can also do a delayed phase, relying on the circulation of contrast if you're unable to pause or slow the flow. So here's an example of a 55-year-old man who was resuscitated after PEA arrest. He had central VA ECMO initiated, and he was sent for a PE study, and a standard injection was performed. So you can see there's dense contrast here in the SVC. But then the next most opacified, structure is actually the venous ECMO cannula. So you can tell here that the contrast is being siphoned off from the venous ECMO cannula that was placed in the right atrium here. And then the next most opacified structures are the inflow cannula and the ascending aorta. So the contrast has bypassed, for the most part, the pulmonary arteries, and you have a non-diagnostic CT pulmonary angiogram. So in this case, the technologists were able to re-inject and reduce the ECMO flow rate and obtain a diagnostic study with the contrast not being siphoned off completely by the ECMO flow. Another thing you can do is inject contrast directly, retrograde into the venous cannula, and here we can see appropriate opacification of the right-sided cardiac structures, including the pulmonary arteries, and there's no central veins that are opacified in the chest. If you're wanting to examine the thoracic aorta or the coronary arteries, you have a challenge in patients on VA ECMO. The ECMO effects can obscure the region of interest due to insufficient vascular attenuation, and you can mistake this for aortic thrombus, dissection, or even aortic occlusion. So what are your solutions? If you have good cardiac output, you can inject through a central line or a peripheral line, and you can decrease that ECMO flow rate or pause it. If you have low native cardiac output, you're gonna have to either maintain or increase your ECMO flow rate and inject contrast directly into the ECMO arterial cannula. Here you have to worry about air bubbles, so this needs to be done by the perfusionist. I prefer to inject through the oxygenator, where you don't have to worry about the air bubbles, and again, the perfusionist will be doing this for you. It's very good to add a delayed phase of image acquisition to rule out a thrombus with certainty, and with central VA ECMO, that's really not much of a problem, because the flow in the aorta is anti-grade. So let's talk about peripheral VA ECMO, and when you're trying to image the aorta and the coronary arteries. So here, this excellent article in Radiographics, published in 2022, shows what's called the watershed area. So if you have the contrast that is injected, we'll go into, you can put it into the oxygenator, but it's going up retrograde into the aorta. So you have two flows here. You have flow coming up retrograde, and that's where the contrast is going, and then you have native flow coming from the heart, pumping out unopacified blood into the ascending aorta, and where these two meet is called the watershed area, and it can be mistaken here for thrombus in the aorta or a dissection. And so you really have to time your contrast and wait until it gets all the way to the aortic root before you start imaging to avoid this pitfall. And here, unfortunately, false diagnosis of potential thrombus in the ascending aorta was made due to poor timing of that contrast or the image acquisition, and there was also potentially some thrombus called in the pulmonary arteries due to poor timing. If the area of interest that you want to image is the cardiac chambers, and that's a pretty common request where you have a risk of thrombus formation due to stasis in the cardiac chambers. So the challenges here are that at first pass, the cardiac chambers may not be opacified if the contrast is directed to the ECMO circuit and bypasses the cardiac structures. And so this also applies to VA ECMO where the contrast is coming in anti-grade, but it's going into the ascending aorta and not into the heart. So solutions are to reduce the flow rate of the ECMO or pause it and inject through a central or peripheral line, or you can save yourself by doing a delayed equilibrium acquisition to detect thrombus and other structural abnormalities. And you can do this at 60 or 120 seconds post-injection. Example here in a young 19-year-old where you have early phase imaging, you've got contrast here in the descending aorta due to retrograde flow of that contrast. You have no pacification of the cardiac structures. You can't rule out thrombus. And a delayed phase acquisition saves the day and proves that there is indeed no thrombus in the cardiac chambers. Another example here of a 45-year-old on peripheral VA ECMO and there is lack of contrast pacification of the right ventricle. Is this true thrombus on early phase imaging and delayed imaging shows that indeed this is true thrombus in the right ventricle. Step five is contrast timing. The factors here are the location of the contrast injection, the time that's required to travel to the vessel of interest. So if you're injecting into the oxygenator, it'll take five to 15 seconds to get to the ascending aorta. And so you don't wanna start your imaging until that time. And you probably don't want, if you're doing triggering, you don't wanna start your monitoring scans until about five to 10 seconds so that you don't irradiate the patient unnecessarily by doing too many monitoring scans. I find that the best way to image these patients is to sit beside the technologist and use a manual trigger. When you see contrast in the vessel or cardiac structure of interest, that's when you start the scan. And then very good to do a late phase, a delayed phase to rule out thrombus completely and also to assess for any perfusion defects. So this is an example of an unfortunate young woman with left main SCAD. And we have early and delayed parenchymal phase. And you can see the extensive hypoperfusion of the anterior wall and septum with sparing of the inferior wall. And she had extensive left anterior descending and left circumflex artery infarction. Finally, scanning technique. Do not use high pitch mode. If you have a dual source scanner, the table motion is, movement is too quick and you are at risk of dislodging these large catheters. And use automated tube voltage selection. Use retrospective gating if needed for the evaluation of the heart function and or the coronaries and iterative reconstruction with metal artifact reduction technique if necessary. So my last example is a 24 year old woman who presented to the emergency department with ventricular tachycardia and elevated troponins. She had a negative CTPA. She was resuscitated and put on peripheral VA ECMO. And we were asked to look at her heart and her coronary arteries to investigate the cause of her VT. We were unable to slow or pause the ECMO flow. So here we can see we used 150 cc's of Vizipake. We did retrospective gating injected into the oxygenator. I sat by the technologist and we used a manual trigger. And we were able to adequately evaluate the coronaries in early and delayed phase imaging. And completely clear the coronaries in this very sick patient. So to summarize, there are six steps to adequately image these patients. The first is, should I do the test? Really look at the clinical question. Step two, patient positioning and table excursion. Remember to practice the table excursion to make sure that your lines have enough slack on them. Contrast injection considerations are important. What vessel or anatomy are you looking to image? Your contrast timing is extremely important. And the scanning technique. So thank you very much for your attention. If you have any questions, please feel free to ask questions online and send me an email. Thanks very much. Thank you. Thank you, Dr. Denny. And safe travels back to Ottawa. Next, we're gonna have Dr. Byung-Wook Choi from South Korea, who's gonna be talking about a tool belt of cardiac devices, X-ray, and CT appearances. Thank you for kind introduction. Today, I'm gonna talk about the appearance of commonly used cardiac devices on X-ray and CT, mainly focusing on X-ray. And cardiac devices refer to a range of medical instruments and implants designed to diagnose, treat, or manage various heart conditions and cardiovascular disorders. These include cardiac implantable electronic devices, such as pacemakers, implantable cardiovascular defibrillators, cardiac resynchronization therapy, and implantable loop recorders. These are designed to control or monitor irregular heartbeats, particularly in individuals with heart rhythm disorders and heart failure. Other cardiac implantable devices for structural heart diseases include heart valve devices, left atrial appendage closure devices, and atrial septal defect closure devices. Additionally, ventricular assistant devices play a pivotal role in supporting ventricular pump function in heart failure patients, which include left and right ventricular assistant devices, intra-aortic volume pumps, and extracorporeal membrane exfusionation. The role of chest X-ray and computed tomography in the context of cardiac device is offering valuable insights at different stages of patient care, such as identification, planning device placement and positioning, evaluation of lead integrity and position, and a complication assessment. Common complications include palpation, infection, thrombus formation, damage to tissue, lung, and vessels, as well as bleeding or leaking. Regarding evaluation of complications associated with cardiac devices, chest X-ray is important to identify potential complications and the CT imaging is very diagnostic. This slide provides an overview of the standardized coding system used to classify pacemakers, detailing their functions and configurations. Each code consists of five positions, with each position representing a specific feature of the pacemaker. The first position refers to the chamber being paced. The options are O for non, A for atrium, V for ventricle, D for dual, meaning for both atrium and ventricle. And the second position specifies the chamber being sensed by the pacemaker, with the same designations. And the third position indicates the pacemaker's response to a sensed event, O for no response, T for triggered response, I for inhibited response, D for dual, and meaning the pacemaker can both trigger and inhibit as needed. The fourth position defines whether the pacemaker has rate modulation capabilities. O means no rate modulation, R means rate modulation is present, allowing the pacemaker to adjust the heart rate based on the patient's activity levels. Finally, the fifth position describes multiple patient capabilities, such as patient multi-sites in a single chamber, O for non, A for atrium, and V for ventricle, and D for dual. Monitoring with a chest radiograph is divided into three distinct phases, early surveillance, maintenance, and intensified monitoring. In the early surveillance phase, the device is evaluated within four to six weeks after placement. This phase ensures that the device is properly positioned and functioning as intended. Following this, we enter the maintenance phase, where the device is routinely evaluated every six to 12 months. This phase focuses on ensuring the continued reliability of the device and addressing any potential issue early. Finally, as the generator battery nears the end of its life cycle, it's typically 10 to 15 years, we initiate an intensified monitoring phase. This involves close and more frequent evaluation to ensure that the device operates effectively until the battery is replaced. By following this structured phase, we optimize the device's performance and enhance patient outcomes over its lifespan. There are representative three types of pacemaker modes. The first one is VVI-type pacemaker. This mode operates ventricular pacing and sensing with inhibitory response. On a chest X-ray, you see the generator and the lead on the RV and read on the RV. It is a left lower and anterior part of the heart here. So pacemaker comprises a generator and battery together in a case, and the single or multiple leads with sensors. The lead in this case, transfers left to innervated vein, superior vena cava, right atrium, and transfers to the tracheal belt to right ventricle. So common indication of this mode is atrial fibrillation and ventricular arrhythmia. Next is AAI pacemaker. It's another popular mode of pacemaker. AAI pacemaker operates atrial facing and sensing with inhibitory response. On a chest X-ray, you see the generator and single lead on RA, the right upper and anterior part of the heart. And this common indication for this mode is a sick sinus syndrome with intact AV conduction. The last common type of pacemaker is DDD pacemaker. DDD pacemaker operates both atrial and ventricular facing and sensing with both inhibition and triggering responses. On the chest X-ray, you see the generator and dual leads in RA and RV. Common indication is a sick sinus syndrome with a failure in AV conduction, second and third degree AV block. The path of the pacemaker leads normally same as before, transverse is the left brachycephalic veins, sophiophenia, right atrium, right ventricle. And next, this is ICD, implantable cardiovascular defibrillator. ICDs are designed to treat life-threatening fast or irregular heart rhythms. They deliver electrical shocks to restore normal heart rhythm. ICD can produce a large amount of electrical energy to defibrillate the heart. ICD is used for primary or secondary prevention of ventricular tachycardia or ventricular fibrillation. The difference from pacemaker is that it has a shock function, so one or two sticks, a shock coil like this. This is a shock coil, and you can differentiate it from pacemaker, a simple pacemaker, because it has a shock coil and thick part of the lead. And another type of the ICD is subcutaneous ICD. It is a special type of ICD. This is a preferred ICD type for pediatric patients or patients for whom intravenous access is not possible. Subcutaneous generator located between the anterior axillary line and the mid-axillary line. The subcutaneous ICD lead with shock coil is located subcutaneous layer of the chest wall parallel to the left side of the sternum and is fixed at the level of xiphoid process. This is, the lead is located at this side. Next, we will move on to cardiac resynchronization therapy, briefly, CRT. CRT devices stimulate both the right and left ventricles to synchronize their contractions. This can be particularly beneficial for individuals with heart failure and conduction abnormalities. CRT devices have leads that are placed in the right atrium, right ventricle, that is the same as pacemakers, but add one more lead to left ventricle. By pacing both ventricles, these devices aim to improve the coordination of the heart's contraction. There are two types of CRT. CRT-D consists of biventricle pacemaker and ICD, and this is more commonly used than the other type, CRT-P, which is a biventricle pacemaker without ICD. There are three leads in the CRT, as I mentioned before. Pacemaker leads are placed in the RA, RV, and LV through the coronary sinus, and the lead to LV through the coronary sinus is located in the posterior part relative to other two leads in lateral view. And this is a CRT-D case, so shock coil is located on RV lead. You can see there is a thick shock coil here. And this is a CRT-P type, same as CRT-D. Three pacemaker leads are placed in the RA, RV, and LV, but it doesn't have a shock coil. As I mentioned before, cardiac implantable electric devices involves various complications, including perforation, infection, thrombus formation, et cetera. One of the most common complications is lead-related, and it includes issues such as lead displacement, fracture, or malfunction. Today, I would like to introduce two complications in this slide. The first one is perforation. As you can see in the left panel, perforation is very, very evident in this patient because the lead extends beyond the normal boundaries of the heart silhouette. And in some cases, CT is necessary to confirm the perforation. In perforation cases, the pacemaker lead may need to be repositioned or removed by simple extraction or surgery to prevent further complications. The second case on the right panel is Tweedler's syndrome. Tweedler's syndrome is rare, but potentially serious complication associated with implanted cardiac devices. It occurs when a patient unconsciously or subconsciously manipulates or rotates the implanted device within the pocket under the skin. This twisting or rotating action can lead to dislodgement of, or dislocation of the device's leads or other components, causing malfunction or ineffective patient defibrillation. As you can see in this case, the lead is winding here in the subcutaneous layer of the chest wall, and the lead tip is dislodged, and the shock coil is abnormally displaced. Management may involve repositioning the leads or in severe cases, surgical intervention to secure the device in place. A leadless pacemaker is a type of cardiac pacemaker that differs from traditional pacemakers. It does not use leads to connect to the heart. Instead, it is a self-contained device that is directly implanted into the heart. Leadless pacemaker has no leads and about three centimeter sized oblong shaped device in the RV apex. In the lateral view, leadless pacemaker located on anterior side of the heart. This shows the process of fixation of the pacemaker in RV trafficulation. The elimination of the leads reduces the risk of lead-related complications, such as lead dislodgement, fracture, or infection. It is a strong advantage of this type of the pacemaker. But it has size and design limitations, which may not be suitable in all patients and in all pacing scenarios. Implantable loop recorder, it is not a pacemaker. But it is a small electronic device that is implanted under the skin of the chest to continuously monitor the heart electrical activity. It is used to diagnose and monitor abnormal heart rhythms and indicated in recurrent syncope, palpitations of unknown etiology, recurrent stroke of unknown cause with a suspected atrial fibrillation, especially over an extended period. Implantable loop recorders generally implanted in the subcutaneous tissue overlying the left pectoralis muscle with a 1 to 2 centimeter incision. The shape is very similar to a leadless pacemaker. So there is a differential point to distinguish those two things. In chest PA views, implantable loop recorder looks similar to a leadless pacemaker. It's like a small USB storage. But it's located in the left pectoralis muscle in the chest wall in the lateral view. In the lateral view, we can easily differentiate the location of the two different pacemakers and these implantable loop recorders. And now, let's briefly review other cardiac devices. TAVR is an alternative to traditional open heart surgery and is less invasive and often recommended for patients with severe aortic valve stenosis who are considered at high or intermediate risk for complications with traditional open heart surgery. TAVR can involve different types of valves, such as balloon expandable valves or self-expanding valves. Self-expanding valves has a longer cylindrical shape than balloon expandable valve. And it is easy to identify TAVR in the chest x-ray if you notice a mesh-structured device at the aortic valve position like this. The MitraClip is a cardiac device used for a procedure called a transcatheter mitral valve repair. It is designed to treat mitral regurgitation. MitraClip is generally considered for patients with mitral regurgitation who may not be suitable candidates for surgical repair. It is often used in individuals with higher surgical risk. This is chest PA for patients who underwent MitraClip procedure. MitraClip is delivered through a catheter and used to clip together the leaflets of the mitral valve at the ventricular side, resulting in reducing the regurgitation. In this patient, there are two clips inserted into the mitral valve. Left atrial appendage closure devices are medical devices designed to prevent blood clots from forming in the left atrial appendage. These devices are primarily used in patients with non-valvular atrial fibrillation who are at risk of a stroke and long-term anticoagulant therapy is not suitable. There are several types of closure devices. In this case, this patient has one of them that is called Watchman device. On chest X-ray, it reflects the shape of Watchman device which is like a spokes of an umbrella, but sometimes it is not easily seen. This is another method to block the entrance left atrial appendage through surgical clipping. You can see a long clip at the left atrial appendage on chest PA and lateral view. Ongoing research is exploring the safety and efficacy of left atrial appendage surgical clipping. This is a very early stage to apply this device. And there are different types of atrial septal defect closure device. Amplature septal occluder is a widely used device with a double-disc design. It is made of a self-expanding 19-ohm wire mesh and is delivered through a catheter. The amplature plug can be used not only for atrial septal defect occlusion but also for left atrial appendage occlusion. The amplature septal occluder can be seen in the middle of the heart on the chest X-ray and partially in the lateral view. This is reprimanded CT views showing amplature device. And next device I would like to introduce is left ventricular OCC device, shortly LVAD. LVADs are used to support patients with end-stage heart failure with reduced ejection fraction. LVAD consists of inflow cannula, impeller, outflow cannula, and drive line connected to a power source and a system controller outside of the patient. But in the chest X-ray, you can see the pump located at the apex and the drive line that connects to the external battery here. But usually the outflow graft is open, not easily visible. This patient has right and left ventricular OCC devices and each ventricular OCC device has the same components. The pump of right VAD is inserted into the right atrium. The pump draws blood from the right atrium and prepares it into the pulmonary artery through an outflow graft. Intra-aortic balloon pump is a temporary circulatory assistive device that works on the principle of cardiac counter-pulsation. The balloon is inflated with gas during diastole and deflates during systole, resulting in increase in coronary blood flow and reduction in left ventricular afterload. Intra-aortic balloon pump catheter with their tips placed too proximally can impede the cerebral blood flow and cause stroke. If the catheter tip placed too distally, it can result in suboptimal myocardial perfusion and potential renal ischemia. The problem is that the only thing that we can see in chest X-ray is the catheter tip that is as a three by four millimeter in diameter and three by four millimeter rectangular metallic density. So the visible catheter tip location is very important as an indicator of a proper positioning of intra-aortic balloon pump, which should be two centimeter above the corona level and two to four centimeter below the aortic arch. This is the last case of my talk. In patients with severe and chronic heart failure, multiple devices are frequently needed and implanted. So in this patient, we can identify LVAD here, ICD with a single lead, and mitral cleft as a small radio-opaque density here. So in conclusion, I would like to emphasize on that the first step in assessing the integrity, positioning, and the potential complications of cardiac devices involves accurately identifying and recognizing their appearance on a chest X-ray, even though CT is very diagnostic, but sometimes it is not appropriate to do CT for simple complication or simple malpositioning. I would like to express my gratitude to Professor Kim, Yoon, and Lee for generally allowing the use of their materials for this talk. Thank you for your attention. All right, thank you. And last talk of the session is my talk. I'm Daniel Vargas from the University of Colorado. And we're going to be talking about complications in left ventricular assist devices. And I decided to do this case-based to go a little bit quicker and show more of these, more images, basically. One of the important things is that there has been a significant change in the use of LVADs. It's actually decreased significantly over the past five to six years. And it's all the result of the revision of the allocation system for a heart transplant. So I know a lot of the patients that would get a left ventricular assist device as a bridge to transplantation are now just waiting for their transplant without the LVAD, because otherwise they get bumped down the list. So this is a better way for them to get it quicker if they don't get the LVAD. Roughly nowadays, we're talking about 3,000 LVADs per year in the US. And the vast majority, if not all of them now, are the HeartMate 3 LVADs. So I try to avoid including any images here of patients with the prior two LVAD types, the HeartMate 2 and the hardware that are now rarely used. So we just saw in the previous talk what an LVAD looks like. And it's a device that gets housed within the chest, within the pericardium, really, in a pocket within the pericardium. And it has a driveline that provides energy. And that's connected to a couple of batteries that the patient carries in a backpack, basically. Two sets of batteries in case one disconnects. There's another one that stays connected. Then you have the pump. This is a circular thing. And it's a magnetically levitated pump. And it has an inflow cannula, which goes within the left ventricle, and the outflow cannula that goes connected to the aorta. On a radiograph, you won't be able to see the outflow cannula. You will see the inflow cannula. And there's also this cover here. This is called the bend relief cover. And it's basically a Gore-Tex material to protect the outflow graft from fracturing at the attachment here. So this is what it looks like on a radiograph we just saw. One, on the prior talk, this is a HarmAce 3 device. It looks like one of those very old phones. And that's how you recognize it. The other ones were a little bit different than this one. This is a driveline that carries energy from the patient's batteries. And notice that you can see the inflow cannula. You can see just a little bit of the outflow cannula here. And the rest of the graft, you won't see it because it's not radiopaque, at least on a radiograph. So this is a case one. And one of the most common reasons for us to image LVADs is to look for LVAD thrombosis. So meaning thrombus within the outflow cannula, inflow cannula, or potentially within the pump itself. Those thrombi, if there's thrombus within the pump itself, we won't be able to see because obviously this is a metallic pump, right? So you can see here the outflow cannula. And this is what I was talking about. This is the bend relief. And it's the Gore-Tex that wraps around the graft. And you can see that there's this hypotenuating material kind of adjacent to the bend relief. And on this axial image, you can see that that hypotenuating material is like crescent shaped. And it might be narrowing a little bit that outflow cannula. So the knee-jerk reaction here is to call this thrombus, but this is not thrombus. This is actually bio-debris. There's a little bit of clot here. There's blood products and other biological debris, but it's not within the cannula. It's between the cannula and the bend relief. And that happens kind of during implantation. And if this material is big enough, is bulky enough, it can actually contribute to some stenosis of the cannula. But the teaching number one here is not to call this thrombus. This is just bio-debris. And I put in parentheses that it's thrombus because it's actually, there's some blood material there, but the term is bio-debris. If there's no hemodynamic significance, they just surveil it. We do see this all the time in these patients. Here's another case, just like the other one. This is a little bit bulkier. But notice it's very smooth, kind of with obtuse margins with the cannula. And again, this is, as you can imagine, that's the wall of the outflow graft, and that's the bend relief, and all this is bio-debris, contributing to a little bit of stenosis here in the proximal aspect of that outflow canal. And this is a little bit bulky. Again, the proximal aspect of that outflow cannula. Very common pitfall. This was another case that we had, and it was interesting because this patient, and this is one of the older devices, this is Harm A2. It was a little bit bulkier. And in fact, this device was not housed within the pericardium. It was actually housed within a pre-peritoneal pocket, so basically in the upper abdomen. But in any case, this is just to show you what we're seeing here. So we have, again, that bio-debris, and kind of along both sides of the cannula. And then you have this other bulky hypotenuating material narrowing the cannula as it bends, right before it attaches to the aorta. And you can see on the axial views that this is a lot more bulky than the bio-debris that I showed you in the previous cases. There's also a little bit of focus of calcification. And with this, our thought was like, this is probably bio-debris. When it becomes irregular like this, you start wondering, is this actually thrombus within the graft itself? And because it was contributing to some narrowing of the cannula, they took this patient back to the OR to open up the cortex covering the bend relief. And then it was actually all bio-debris outside it, and the cannula re-expanded after that. So again, even though it looks a little bit bulkier and irregular, this was also considered bio-debris. And I guess, again, one of those things that helps us call that as bio-debris is those smooth margins and obtuse angles with the cannula. Here's another case. This patient had increase in LDH and a blunted RAM study. I'm gonna tell you what a RAM study is in a second. With the CTA looking for thrombus, and lo and behold, there is complete non-opacification of the outflow cannula here. So this is obviously a thrombosed outflow cannula. It's actually not very common, because obviously these patients are anti-coagulated. Why do these patients get thrombosis? There's flow-induced thrombogenicity of the device itself. There's an inflammatory and hypersensitivity response to having this device there, too. Those are the main causes for thrombosis. It can be a pretty devastating complication. And it can be, these patients can be either asymptomatic all the way to the end of the spectrum of a cardiogenic shock. Very useful to do a radiograph as a first means of evaluation of these patients, because they'll have pulmonary edema, and that will be one of the first things that we'll see. Obviously, these patients will have a change in functional capacity. They'll have power spikes in the LVAD pump, increase in LDH with quite a high sensitivity and specificity, and obviously increase in BMP, et cetera. So what is a REM study? It's a REM study, it's basically an echocardiogram that is repeated at different pump speeds. So this pump, if you can imagine, this pump is sucking blood from the left ventricle. If you increase the speed of the pump, it's gonna suck more blood out of the ventricle. So your left ventricular and diastolic diameter or volume should decrease. So that's what we're looking for. We increase the speed. We, the cardiologists, when they're doing this, and the expectation is for the left ventricular and diastolic diameter to decrease. The degree of mitral regurgitation also to decrease as more blood is sucked out of the ventricle through the pump. And also you would expect the AV valve, the aortic valve to open as well. I'm sorry, to not open as commonly because you're sucking through the device. So when they have a blunted or absent reduction of that left ventricular and diastolic diameter with that increase in the RPM, then you can imply that there's perhaps an obstruction to the flow in the device. And we're not gonna know where it is. And that's why we do the CT. Hopefully we can see a thrombus within the cannula. Here's another companion case. This is also a patient with some thrombus. This is, again, see the angles here are not as smooth as they are in the area of just bio-debris. This is a little bit of thrombus within the cannula. And this patient had embolic events, which is obviously one of the risks. One of the other things that we see in these patients is that the patients with LVAD, because the blood is being sucked from the ventricle, there's no pressure gradient generated across the aortic valve to open the aortic valve routinely with every heartbeat. So the aortic valve might open every two, every three, every four heartbeats. And because you have stasis of blood with the aortic valve not opening, these patients can develop thrombus. You can see some thrombus here in the sinus of valsalva just due to a stasis over that aortic valve leaflet, as you see here. And you can see this patient has an LVAD. And as expected in the CTA, you can see beautifully that as you expect, that the contrast of the pacified blood is being sucked through the inflow cannula, and there's no pacification of the ascending aorta. Thrombotic embolic complications are very common. If you have thrombus built up within the device and related to the thermogenicity of this device. And this is a patient with embolic events, obviously a large stroke, some infarcts in the spleen as well, related to the LVAD. All right, this is a very simple case. The patient with an LVAD, as you can see here. And again, you can see there's a little bit of biodebris, but obviously the big finding here is this large pericardial fusion with high attenuation implying blood product. So there's a big hemopericardium. And this is just to talk about the hemorrhagic complications of LVADs. They're very common. In part because of an acquired von Willebrand syndrome, because of von Willebrand molecules will be destroyed by the pump. These patients are anticoagulated. They have antiplatelet therapy. There's impaired platelet aggregation because, apologies for the typo there, because of the pump. And interestingly, that continuous flow physiology without the systole and diastole promotes angiodysplasia. So the most common type of bleed in these patients is gonna be GI bleed related to AVMs in the bowel. You can see the incidence of GI bleed in multiple studies, pretty high, between 14 and 35% of the patients. And a lot of these patients related to AVMs in the bowel. And that's thought to be a result of that continuous flow physiology. Hemorrhagic stroke, very common. Other bleedings are also fairly common. Retroperitoneal bleeds, like the one I'm showing you here. Pump pocket hematoma, like the one I showed you with the hemoperic, or if the pump is in the preperitoneal region. And obviously, if the pump is within the pericardium, then you have a hemopericardium. Here's another picture with an LVAD, unfortunately with a large hemorrhagic stroke in the right frontal lobe. And again, another case of hemopericardium, very large hemopericardium. Obviously, if we have contrast here, if we had contrast, one of the things that you need to do when you have a large hemopericardium such as this one is look for evidence of cardiac tamponade. You don't wanna have a patient that's on an LVAD for left heart failure or also developing right heart failure due to a tamponade physiology. And the other big reason why we image these patients is for LVAD infection. They're fairly common, although with the newer LVADs, infections have become less and less common. And this was a patient that we did suspected infection. You can see there's a large fluid collection throughout the pump pocket surrounding the outflow cannula. There's also some fluid extending through the median sternotomy and through the chest walls of tissues. Obviously, this is an infection going on, just a multiloculated abscess in this case. So how do we define LVAD infection? There's a couple of definitions. There's the percutaneous driveline infection. That is an infection that follows a driveline. And it's very common. Why? Because the driveline is percutaneous. So you're gonna have infection entering from the skin through the driveline. And the infection can be superficial if it's superficial to the muscular fascia or deep if it involves the muscles. Then you're gonna have a pocket infection like the one I just showed you, or a pump infection per se, which would be akin to an endocarditis. We won't be able to see that. That's just gonna be a lab and basically a clinical diagnosis of endocarditis in a patient with an LVAD. And those are a very high mortality, as you can imagine. So the driveline infection, as I said, very common in some series up to 50% of the patients. A pocket infection, luckily, is not as common. Obviously, the most common infectious agents are gonna be staph, MRSA, some pseudomonas as well. And this is a driveline infection. You're gonna see some stranding and some non-organized fluid here along the driveline. And sometimes you'll see an actual abscess here, but this would be superficial. It's superficial to that fascia. Compared to this one, this is a deeper infection. It's not only involving the musculature. There's a myositis that you can clearly see here, but it's also extending deeper into the pocket with some stranding and, again, non-organized fluid in the region of the pocket. And again, just for another example, there's an abscess following that outflow cannula in a patient with an LVAD infection, a pocket infection here. This is another case. This was right after implantation, and for some reason, we did a CTE chest without contrast, and the patient had air in the cannula, and people were freaking out because they thought that this was just gas-forming organisms related to the LVAD. This is just postoperative air that just got in, and it's basically, again, between the cortex, ven relief, and the cannula. So, this would be where the bio-debris forms, so just air, innocuous air, and you can see that there's also a little bit of air in the mediastinum, too. And I think this is the last case I have. This patient presented with power spikes of the pump, concern for thrombosis. You can see that there's an LVAD in place. This was one of the older types, HARM-A2, and again, bio-debris, very common here, but not thrombus. We can see the inflow cannula, and there's not, whatever we can see of this shows that there's no thrombus within it, but it's obviously in a position that is not expected, right? This thing is directed towards the apical portion of the interventricular septum, and it shouldn't. The LVAD cannula should be directed towards the mitral valve, parallel to the septum, and if this video runs, here you can see how it's kind of sucking in the interventricular septum, and hence this patient is getting power spikes from the pump, because it's not able to withdraw blood from the ventricle as well as it should. That's a malposition LVAD. As I said, the inflow cannula has to be perpendicular to the septum, directed towards the mitral valve. The outflow cannula, if it's too short, it can compress the RV as it goes up towards the aorta. If it's too long, it can kink, and that kink can cause a flow obstruction too. Here's a couple of examples, oops, went the wrong way. Couple of examples of malposition LVADs. Even though there's no obstruction, this LVAD should not be directed towards the anterior wall. Same with this one, different patients. These should be directed towards the mitral valve. So slightly malpositioned in this case. And this is an outflow cannula that is kinked. It was a little bit redundant, and you can see this kink here, this bend, and that can contribute to flow obstruction. So if there's a gradient across this, that, and the patient is getting power spikes from the pump, they might need to go in and fix that surgically. Here's another case of a more severe kink of the outflow cannula too. And with that, I do want to acknowledge Garia Shroff, Daniel Ocasiones, and Phil Young for sharing a few cases. We published this several years ago now. You can find this, and there's a lot of these examples and others in this paper. And with that, I thank you very much for your attention, for staying. You guys are champions. 4 p.m. last day of RSNA, so hopefully we'll see you next year. Thank you.

ECMO

CT imaging

cardiac devices

Venovenous ECMO

Venoarterial ECMO

contrast dilution

catheter displacement

pacemakers

LVAD complications

thrombus detection