Hello, my name is Brandon Wiley. I'm the Chief of Cardiology at Los Angeles General Medical Center and Associate Professor of Medicine at the Keck School of Medicine at the University of Southern California. And today, I'm going to present cardiovascular surgical complications and mechanical devices. I have no disclosures. The objectives are straightforward. We're going to discuss the differential diagnosis and management for the decompensated patient after cardiac surgery, and we're going to review the function, management, and troubleshooting of pacemakers, ventricular assist devices, intra-aortic balloon pumps, and extracorporeal membrane oxygenators. Let's start the talk discussing post-cardiac surgery complications that we see in the ICU. These include vasoplegia, low cardiac output syndrome, right ventricular failure, cardiac tamponade, bleeding, arrhythmia, myocardial infarction, and cardiac arrest. Vasoplegic syndrome is a common complication of cardiac surgery with a high mortality. It presents within 24 hours and is defined as severe hypotension requiring vasopressor therapy in the setting of normal or elevated cardiac output and low systemic vascular resistance. Important risk factors include patients that were taking ACE inhibitors or ARBs prior to surgery and those surgeries that had prolonged clamp times or that were more complex in nature, such as redo surgeries or multivalve operations. The proposed pathogenesis is complex. There is systemic inflammation, ischemia and reperfusion injury, and exposure to the synthetic surfaces of the cardiopulmonary bypass circuit that drive increased oxygen-free radicals, nitric oxide, and increased levels of pro-inflammatory cytokines and other vasoactive substances. All these factors drive the development of a systemic inflammatory response syndrome. Additionally, there's a reduction in arginine vasopressin, which contributes to excess nitric oxide and vasodilatation. And on top of that, there's inducible nitric oxide synthetase, which produces nitric oxide and increases cyclic GMP and creates even more vasodilatation. The treatment of vasoplegic syndrome is mostly supportive and includes goal-directed fluid resuscitation and the use of vasopressors to maintain a mean arterial pressure greater than 65. What is the first-line vasopressor agent? Well, it's probably vasopressin. In the VANCS trial, vasopressin versus norepinephrine, vasopressin reduced acute renal failure and atrial fibrillation in the post-operative setting, but there was no difference in mortality. Norepinephrine is the next-line vasopressor, and it can be used in combination with vasopressin. Really, dopamine should be avoided due to risk of arrhythmias. There are adjunct therapies that can be used for refractory cases, but data is limited. Methylene blue is a nitric oxide synthetase inhibitor. It is given as a bolus, but the duration of action is fairly short, about 40 minutes. We need to be cautious in those patients with G6PD deficiency, as it can induce hemolysis. Methylene blue is also known to precipitate serotonin syndrome. Serotensin 2, in a post-hoc analysis of 16 post-cardiotomy patients, was shown to reduce requirements for other vasopressors, so this is a medication that could be used as well. And then hydroxocobalamin, or Cyanakit, which is a B12 precursor, that's a direct nitric oxide scavenger, and its use can reduce norepinephrine requirements. Although there really is a heterogeneous response, there are responders and non-responders, and this really may work better as a continuous infusion. And finally, corticosteroids. Not a lot of data here, but they're probably safe, given that they have been studied in high doses intraoperatively, and there wasn't found to be significant complications. Another complication of cardiac surgery that is common and associated with high morbidity and mortality is low cardiac output syndrome. Now, low cardiac output syndrome has been variably defined in the literature, but in principle it is when the cardiac index is less than 2, the systolic blood pressure is less than 90, and there are signs of tissue hypoperfusion, such as oliguria, elevated lactate, cold extremities, or poor capillary refill. Invariably, these patients require inotropic agents or mechanical circulatory support. Now there are preoperative risk factors for the development of low cardiac output syndrome. These include the presence of cardiogenic shock or heart failure prior to surgery, or those patients that are undergoing redo cardiac surgery. In addition, there are factors associated with the surgery itself. Important ones to consider are myocardial stunning or ischemia that is a result of poor myocardial protection. Additionally, there are reversible causes such as arrhythmias or tamponade that need to be dealt with. Ultimately, all these factors lead to left ventricular and or right ventricular systolic and diastolic dysfunction. Now the treatment of low cardiac output syndromes should start with a comprehensive evaluation for reversible causes. And it is important to monitor these patients hemodynamically, ideally with a pulmonary artery catheter, as intercardiac filling pressures and systemic vascular resistance assessment is crucial really to make decisions regarding therapeutic interventions, and then to monitor the efficacy of those interventions. We need to correct any sort of electrolyte derangements, hyperkalemia and hypocalcemia are really important to correct. And then we need to regulate acid base status and optimize oxygenation and ventilation. Importantly, heart rate optimization is critical. We need to maintain AV synchrony whenever possible because that helps improve cardiac output. We need to treat arrhythmias aggressively, either with cardioversion or antiarrhythmic medications. And then finally, the choice of vasoactive medications really should be guided by the hemodynamic data. And we need to have a low threshold to move forward quickly with mechanical circulatory support if perfusion does not improve with medical therapy. Another complication of cardiac surgery is right ventricular failure that requires prolonged inotropic support or mechanical circulatory support. This is most relevant in those patients that undergo durable left ventricular assist device placement, but we see that is not uncommon in all cardiac surgery patients. It is associated with increased risk of morbidity and mortality. Importantly, there are preexisting risk factors that we need to be wary of, and we really need an echocardiogram. We're looking for RV dysfunction at baseline. And then invasive hemodynamics are very helpful, particularly, again, in those patients that are undergoing LVAD placement. In the postoperative period, we can look at those operative risk factors, particularly, for example, patients that required long pump times, those patients that suffered RV ischemia from air emboli. And then we need to be careful with excessive volume loading in patients because that can exacerbate severe tricuspid regurgitation, and again, some RV dysfunction in that setting. And then, of course, complex valve surgeries or those patients that had a protamine reaction at the end of their bypass run. The physiology of RV failure is driven by multiple factors, including arrhythmias, pressure overload, RV volume overload, and decreased RV contractility. All these lead to RV systolic impairment, which then leads to RV dilation. In the setting of RV dilation, the tricuspid annulus is enlarged, and there is worsening tricuspid regurgitation, which creates this cycle. Again, when the RV dilates, the interventricular septum is pushed over to the LV. This puts the LV at a mechanical disadvantage, reduces LV filling, and reduces systolic function and stroke volume or cardiac output, which then drives RV ischemia, and you create, again, this cycle. Importantly, we need to monitor invasive hemodynamics in this setting, and we have some markers here. I'll point out the PAPI or pulmonary artery pulsatility index of less than 1.5 is an important marker. And then we want to monitor everything with echocardiography. This is a way for us to quickly see RV dysfunction and the severe tricuspid regurgitation that is present. The management of RV failure focuses on reducing the causes of RV dysfunction. That is, we want to reduce the volume overload that the RV is experiencing. We want to reduce the afterload on the right ventricle. We want to improve RV contractility. We want to optimize the rate and rhythm. We want to ensure AV synchrony and treat any arrhythmias. And then finally, we want to ensure adequate perfusion to the right ventricle, usually targeting a slightly higher mean arterial oppression, 70 to 85 in those patients that are post LVAD. If we're moving to mechanical circulatory support to treat RV failure after cardiac surgery, the most important issue to understand is whether or not we're dealing with isolated RV failure or biventricular failure, as this will guide our device choice. Our next complication of cardiac surgery is a perioperative myocardial infarction. This is termed a type 5 myocardial infarction by the universal definition of MI in 2018. And really we're looking at troponin elevation within 48 hours after cardiac surgery. And with that troponin elevation, we need to have evidence of ischemia, and that is either on the ECG, QAs, or a new left bundle branch block, for example, or by angiographic evidence in the cath lab, or finally by imaging evidence such as echocardiography showing a decrement in left ventricular ejection fracture and new wall motion abnormalities. If a perioperative MI is suspected, what can be done about it? Well, the key issue is to identify those patients with actionable lesions, as this group may benefit from urgent revascularization. Notably, saphenous vein graft occlusion is not uncommon in the postoperative period, and it is associated with increased mortality. We have to be on the lookout for those patients with recurrent ventricular arrhythmias, as this could be a sign of ongoing underlying ischemia. The ECG can be unreliable, and really we need to move towards coronary angiography as that is the gold standard. An emergent percutaneous intervention is better than redo cabbage for perioperative mortality. Moving on to bleeding. All patients bleed after cardiac surgery, and bleeding is associated with an increased morbidity and mortality. How do we define significant bleeding after cardiac surgery? The universal definition of perioperative bleeding in adult cardiac surgery introduced a classification scheme that incorporates chest tube output and the amount of blood products transfused. Mortality increases with increasing classes in bleeding based on this criteria. And we want to treat bleeding and reverse coagulopathy in the ICU as re-exploration for bleeding in the OR is associated with increased mortality. There are multiple factors that increase bleeding after cardiac surgery. An important one are antiplatelet therapies. So for example, P2Y12 inhibitors such as clopidogrel, ticagrelor, and prasugrel increase the risk of bleeding after surgery. What is listed here is the number of days prior to cardiac surgery that these medications should be stopped based on the guidelines. Other causes to think about are prolonged surgeries or complex surgeries. And then of course, we need to deal with the residual effects of heparin and fibrinolysis from the cardiopulmonary bypass circuit. The STS clinical practice guidelines for patient blood management was updated in 2021 and included data from more than 8,000 patients in randomized control trials. The data pointed to targeting a restrictive transfusion policy, that is to say a trigger of around 7 to 8 versus a more liberal transfusion policy. So for example, a trigger around 8 to 10. Because what they found was with a restrictive transfusion policy, it re-reduced the number of transfusions without increasing the risk of mortality or morbidity. And then in order to really target our transfusions and target our ability to manage the coagulopathy, we should be using thromboelastometry or thromboelastography so that we understand what coagulation products we need to give the patient to reverse the coagulopathy and reduce the risk of return to the OR. And then of course, we need to work with the surgeons in those severe cases, such as when there's a sudden increase in drain output or we suspect tamponade. Now cardiac tamponade, this is important. This is something that we really need to be aware of in the ICU because if we can recognize this early and we can treat it, we can really get and see full recovery of the patient. Now it's important to note that you can have or develop tamponade early in the perioperative course. So that is less than 7 days out. Or late in the perioperative course, that is greater than 7 days out. And if patients develop tamponade greater than 7 days out, there is higher mortality. The diagnosis can be made using invasive hemodynamics. You're looking for an elevated CVP, loss of the Y descent, low cardiac output, for example. But it is important to understand that you can have cardiac tamponade with low CVP. And that's in the setting of regional or focal compression. Also those patients on mechanical ventilation, if they have an arterial line in place, for example, you'll see reverse pulses paradoxes. Or you may see blunting of pulses paradoxes with mechanical ventilation. And you may not see much pulses paradoxes in cases of focal or regional tamponade. Now echo is incredibly important here because we can make the diagnosis not only of the effusion but of the hemodynamic compromise that exists. But importantly, we should remember that transthoracic echocardiography has limited sensitivity in the postoperative setting due to surgical bandages, for example. So if you have a high pretest probability of tamponade, we really need to move forward to a TEE because that is our diagnostic modality of choice. This is a bedside ultrasound that demonstrates focal tamponade of the right atrium. We have coagulum and hematoma here that's compressing the right atrium. We see when we put color, there's acceleration of flow across the tricuspid valve and the right ventricle is underfilled. This is just an example of how transesophageal echocardiography can be so helpful in cases of tamponade after cardiac surgery. This is a patient that we took back to the OR for bleeding. And incidentally, we found a pericardial hematoma right here. Luckily, this hematoma was not compressing the right ventricular outflow tract, but we could see if it did grow, it would cause focal compression. So again, an example of how transesophageal echocardiography can be very helpful due to its improved resolution and ability to recognize pericardial hematomas that can't be seen with transthoracic echo. Let's move to a question. We have a 62-year-old woman who has type 2 diabetes, hypertension, and unstable angina. She undergoes CABG, an aortic valve replacement, and now she's post-op day zero. You're en route to evaluate her at the bedside for hypotension, and she arrests. You diagnose pulseless ventricular tachycardia. According to the most current STS guidelines, you should direct your ICU team to do the following. Start chest compressions, perform cardioversion, and then follow ACLS procedures. Start chest compressions, perform cardioversion, follow ACLS, and prepare for re-sternotomy. Perform cardioversion once, start chest compressions, and continue ACLS. Perform cardioversion once, start chest compressions, continue ACLS, and prepare for re-sternotomy. Or perform cardioversion three times, prepare for re-sternotomy, and start chest compressions. Well, if you chose perform cardioversion three times and then prepare for re-sternotomy, you'd be correct. Cardiac arrest after cardiac surgery is typically due to ventricular fibrillation, and because of that, the STS guidelines that we see here recommend defibrillation with at least three attempts before starting CPR. We want to avoid epinephrine because if we get severe hypertension after resuscitation, that can lead to increased bleeding. And then we wanna move to early re-sternotomy in these patients, which is performed typically in the ICU. Our last diagnosis for hemodynamic instability in cardiac surgery is arrhythmia. And supraventricular tachyarrhythmia is really common after cardiac surgery. The most common form of supraventricular tachyarrhythmia is atrial fibrillation or flutter. It's typically associated with valvular surgery. Beta blockers, such as metoprol, reduce the incidence of post-op AFib. However, beta blockers are negative inotropes. And therefore, we wanna avoid those in cases where we need inotropic support. Amiodarone is very effective, both for reducing the incidence of post-op AFib and for pharmacologic cardioversion. It's also less of a negative inotrope compared to metoprol. And that's what's really useful. It's really useful in cases where we require inotropic therapy or mechanical circulatory support. However, we need to be aware that amiodarone does have potential side effects, such as pulmonary, thyroid, and liver toxicity. Sustained ventricular tachyarrhythmia is rare, but when it occurs, we really need to think about ischemia. And then bradyarrhythmias are common. We see about one in five patients with a bradyarrhythmia. The AV node is located near the tricuspid, mitral, and aortic annuli. And therefore, there's more risk for development of AV conduction abnormalities after valvular surgery. Luckily, bradyarrhythmias are usually transient, and only about 2% to 4% of patients require permanent pacing. So let's move forward to mechanical devices. Let's talk about epicardial pacing, and then we'll talk about mechanical circulatory support. So epicardial pacing wires are placed after cardiac surgery in the OR, and they are used to manage conduction abnormalities and improve hemodynamics. These are typically placed on the right ventricle and at times the right atrium. We're able to treat sinus node dysfunction and AV conduction defects. We can utilize leads on the atrium and ventricle to establish AV synchrony. And then we can overdrive tachyarrhythmias to terminate them. There is some data around prevention of postoperative atrial arrhythmias, but you need biatrial pacing for that. Now, temporary epicardial pacemakers utilize the same nomenclature as permanent pacemakers. Pacing modes are defined as the chamber paced, the chamber sensed, and the response to sensing. This slide lists the most common modes that are used in the ICU and their indications. The presence of atrial and ventricular pacing wires allows a practitioner to switch between modes to manage different dysrhythmias and to optimize hemodynamics through the maintenance of AV. In those patients that are being paced, the first step is the assessment of the underlying rhythm. The best method to evaluate the underlying rhythm is to gradually reduce the pacing rate to reveal the intrinsic rhythm. Reducing the pacing output or unplugging the pacing leads can be really dangerous in patients without intrinsic rhythms. The next step is to assess the capture threshold. This is incredibly important in patients that may be pacer dependent. Increasing capture threshold over time, that is more voltage being required to capture the myocardium over time, may be an early sign of pacemaker failure. Pacing creates inflammation and fibrosis at the lead site, and that may increase over time the amount of energy needed to capture that part of the myocardium. To check the sensitivity, the pacemaker should be set to VVI, AAI, or DDD with a rate just below the intrinsic rate. The sensitivity is then reduced, that is to say the millivolts are increased until the pacer is pacing asynchronously. The sensitivity should then be increased or the millivolts decreased until the generator's sensor begins detecting those intrinsic depolarization and inhibits pacing. The pacer can be set to half the sensing threshold. With over-sensing, the pacemaker is too sensitive. That is to say that it will misrecognize pacing spikes or other electrical signals as endogenous depolarizations and thus be inappropriately inhibited. With under-sensing, the pacemaker is not sensitive enough to recognize endogenous depolarizations and therefore will inappropriately pace, causing output to synchrony. Output to synchrony can be a real problem for patients as it can lead to decrease in cardiac output, increased cardiac work, or even R on T phenomenon. Let's talk about troubleshooting pacemakers. Failure to pace is the first issue that we need to deal with, and that's when there's an absence of a pacing spike. When you're faced with this situation, what you can do is switch the pacer to an asynchronous mode such as AOO or VOO. If there is pacing when the device is switched to an asynchronous mode, then you've ruled out the presence of lead failure, an unstable connection, or insufficient power. Now, crosstalk occurs when you have a dual pacing setup. So, for example, DDD, and in that setting, the atrial pacing spike can be sensed by the ventricular lead and interpreted as an intrinsic ventricular beat. This will inhibit the ventricular lead from pacing. The opposite can occur when the atrial lead senses the ventricular pacing spike as an intrinsic atrial beat. Crosstalk can be managed by reducing the sensitivity or by simply reducing the power output, which reduces the amplitude of the pacing spike. Now, failure to capture occurs when there's a pacing spike present, but you're not capturing the myocardium. And this can due to a problem at the site of the pacer lead implantation. For example, the development of inflammation or fibrosis due to pacing, or it can be due to a more global issue such as electrolyte derangement, for example, hyperkalemia or acidosis, which raises the threshold of the myocardium for depolarization. Pacemaker-mediated tachycardia is an issue with over-sensing that can only occur with dual chamber pacing. Pacemaker-mediated tachycardia occurs when an atrial lead senses a ventricular pacing spike as an intrinsic atrial beat, and thus there is a ventricular impulse that is generated. This over-sensing can lead to an endless loop tachycardia. To prevent this tachyarrhythmia or to break the tachyarrhythmia, you can switch the device to VVI, and to prevent it, you can increase the atrial blanking period. Pacemaker-mediated tachycardia can also occur through retrograde conduction between the ventricle and the atrium, either through the AV node itself or down an accessory pathway. This can be prevented by adjusting the post-ventricular atrial refractory period, or PVARP. If this occurs, you can switch the device to VVI or DVI to break the tachycardia. We talked about epicardial pacing. Now let's move to mechanical circulatory support. Let's start with an intra-aortic balloon pump, which is an incredibly common form of percutaneous mechanical circulatory support. The intra-aortic balloon pump inflates during diastole, and then rapidly deflates during systole, creating counterpulsation. Physiologically, when the balloon inflates, it increases diastolic pressure, and you get improved coronary perfusion pressure, and then it rapidly deflates during systole, decreasing afterload, and with that myocardial oxygen demand. Overall, this results in a rather modest increase in cardiac output, which is around 0.5 to one meters per minute. Now the device, it times off of a trigger, and what you want to happen is for the balloon to open right at the dichrotic notch here, creating this augmented diastolic peak pressure, and then deflate right prior to systole here. Typically, the trigger that's used is the R-wave on the ECG, but if you have a poor ECG, you can time it off of systolic pressure. Some key indications for an intra-aortic balloon pump to think about are the use in cardiac surgery patients who have cardiogenic shock after bypass. We also use it in patients with unstable ischemic disease, so those patients with refractory angina, multivessal disease, for example, or those patients with severe left-main stenosis who are awaiting surgery. It's very helpful in patients with mechanical complications of MI, such as those with severe mitral regurgitation or ventricular septal defects, and then we're going to talk about its use with VA ECMO to assist in LV unloading. Some important contraindications to think about significant aortic regurgitation, because that will get worse with an intra-aortic balloon pump, those patients with aortic dissection, and then those patients with really severe peripheral vascular disease, because they're at risk for ischemia of their limb. So let's talk about timing. With optimized timing, we want to see that the augmented peak diastolic pressure right here is greater than the non-augmented systolic peak pressure, and then with deflation, what we're looking for is this diastolic runoff here. So we want to see the assisted end diastolic pressure. That should be less than the unassisted diastolic pressure, ideally by about 15 millimeters of mercury, and then the assisted systolic pressure here should be less than the unassisted systolic pressure here, at least by five millimeters of mercury, and that's demonstrating that reduction in afterload that we like, which improves cardiac output. Now let's recognize some timing errors using the hemodynamic tracings. Let's start here in the upper left. This is a problematic issue. This is early inflation, and what we see here is the balloon inflating before the dicrotic notch. That's causing premature closure of the aortic valve, which decreases cardiac output and increases cardiac work. We want to shift the trigger a little later here. Down here, we see late inflation. This occurs when a balloon inflates after diastole has already begun. We can see that the dicrotic notch is here. In this setting, we are not getting effective counterpulsation. We don't get that increase in the diastolic pressure, so we're losing that improved coronary perfusion pressure. However, we don't worsen our cardiac output quite as much as with early inflation. Early deflation is occurring here, and this is important because what we see is that return of the diastolic aortic pressure towards normal, and because of that, we're not getting that reduction in afterload, and we see that our assisted systolic beat is about the same as our unassisted systolic beat. And then finally, we see here late deflation. Late deflation is an important issue to recognize. With this, the balloon is staying inflated into systole, which dramatically increases afterload and worsens cardiac output. A couple of key management points and complications. The first is that tachycardia, particularly irregular rhythms such as atrial fibrillation with rapid ventricular response, reduce the hemodynamic benefit of the intraordinary balloon pump as the balloon just cannot time well, and so it can be helpful in those patients to reduce their ventricular rate with pharmacologic agents. We really wanna track where the balloon pump is. We use the radio-opaque distal marker. It should be about one to two centimeters distal to the left subclavian at the level of the carina. Daily chest X-rays are helpful to make sure that we're not seeing migration of the device. And then anticoagulation is recommended when there's an intraordinary balloon pump. However, when the balloon is at one-to-one, augmented to non-augmented, you may not require anticoagulation. If the patient is high-bleeding risk, you could probably hold it. However, once you drop the ratio of augmented to non-augmented beats to one to two or one to three, the patient should be on systemic anticoagulation, and you should never leave the balloon pump in the patient uninflated unless the patient is anticoagulated. And we need to monitor for limb ischemia daily. And then we need to be aware for other possible issues such as catheter kinking, balloon rupture, where you'll see actual blood within the catheter itself. And then intraordinary balloon pumps are associated with some amount of hemolysis and thrombocytopenia. So let's move from intraordinary balloon pumps to more advanced mechanical circulatory support devices. Let's talk about microaxial mechanical circulatory support devices or impella devices. These are transvalvular univentricular support devices that can either be used for left ventricular or right ventricular support. They come in several different flavors. The CP, which provides about 4.3 liters of flow per minute. The 55, which can be placed directly into the ascending aorta or through an axillary cut down, which provides about six liters of flow. And then the RP impella, which is for RV support and can provide about four liters of flow. Now, left-sided devices, contraindications to think about is the presence of LV thrombus. You can't insert these devices across a mechanical aortic valve if there's moderate and severe aortic regurgitation. Critical AS or really severe peripheral arterial disease in patients where you're placing this peripherally can lead to menischemia. With an RP impella, again, you want to avoid placing it in patients that have RV thrombus or severe pulmonary valve stenosis or regurgitation. And again, you're not going to advance it across mechanical valves. And then these devices are associated with increased risk of bleeding and hemolysis as well as thrombocytopenia. And as mentioned, menischemia is an important issue and needs to be monitored over time. This is what an impella CP looks like in the left ventricle. We see this ring down artifact here with the echocardiogram, that's the inlet. That should be positioned in the mid ventricle and the distance from the inlet here to the aortic valve with the CP should be about 3.5 centimeters. The tip of the impella should be free of any of the subvalve apparatus of the mitral valve. And the outflow as seen with color here should be in the ascending aorta. Now the impella 5-5 does not have a pigtail tip and it's a bigger device. And we can see that the inlet here is actually a little posterior. There's a ring down here. The distance from the inlet here to the aortic valve should be about five centimeters. Now it's important to assess these devices with echo. Echo is really the key imaging modality and it should be used anytime that the device needs to be repositioned. And it should be part of the daily assessment of the device at the bedside. There's the inlet here with the ring down and that's the distance you're measuring from the annulus to the inlet. And here again, that distance that you're measuring from the annulus to the inlet. So let's talk a little bit about the impella management. Again, we wanna use echo to guide all device manipulation. We control the speed of the impella device with the console here. And it ranges from zero, which is the device off to nine, which is maximum. Typically we operate in the P5 to P8 range. And really we wanna have the lowest P level or the lowest speed level that provides adequate support because the higher your speed, the more likely you are to have issues with hemolysis, thrombocytopenia and suction. And we only wanna turn the device down to P1 if we're removing the device from the ventricle. Now, the device gives us a placement signal here. This is a pressure tracing of the ascending aorta. And then it calculates a left ventricular pressure as well. And we can use these two pressures to assess for position of the device and or suction. And we can see that if the AO tracing here turns into an LV tracing like it is here, that's telling us that the device has migrated into the left ventricle. The motor current is depicted here and we should see some pulsatility. And if we see increases in the motor current over time without changes in flow, that can suggest an issue with the pump such as thrombosis. Now, the pump requires a purge solution. What's recommended is heparin through purge. If the patient has bleeding issues or has developed HIT, then you can use a sodium bicarb solution. And then the impellos are recommended to be run with an anti-10A level of around 0.2 to 0.4. You can get there with that purge solution of heparin. But if you aren't there, then you may have to run some heparin peripherally. This is an echocardiogram showing shallow placement of an impella. We see here the impella is being actually ejected into the ascending aorta with each beat. And again, we're comparing that to our normal placement of the impella here. The inflow here, we see that ring down, is free of the subvalvular apparatus of the mitral valve. And it is at least about 3 and 1⁄2 centimeters from the aortic annulus. Now, impella management, things we want to think about, one is a suction event. And you can have acute suction or continuous suction. You can identify continuous suction if you see a decoupling or an uncoupling of the AO tracing here of the device and the LV pressure tracing that's calculated here. See how these have uncoupled from each other. The other thing we're seeing here is a very negative LV diastolic pressure. And that can be a tip-off as well that we have suction. And then we see here quite a low diastolic flow and a low flow in general. All that can tip us off to that we're having a continuous suction. You can also just have diastolic suction itself. And what we see here is, although there has not been an uncoupling of the signals, we see that there is still that very low LV calculated diastolic pressure. And we see this abnormal diastolic pressure change here in diastole. And we see a very low diastolic flow here. Now, how do we deal with suction? Well, we reduce the speed. That's the first thing you want to do. The next thing you want to do is evaluate the positioning of the device with echo because the device can migrate. Other things we want to do is ensure adequate preload. The 5-5 device requires at least a CVP of around 10. And then the other thing that can lead to suction is RV dysfunction. So bowing of the interventricular septum into the left ventricle can create a suction event. The other issue we want to think about other than suction is hemolysis, which we see with this device. Really, if you're seeing hemolysis, which you should be trending daily LDH to sort of track for that, you really want to, again, image the device with echo and make sure that you're not up against the wall and the device is in a good position. You can turn down the speed as well, drop your P rate, and see if that doesn't help to reduce the amount of hemolysis that you're seeing. This is an echo example of acute suction event. This is an impella CP. That'll be a suck down on the impella. So we reduce the speed really quickly, and we see re-expansion of the left ventricle with the impella here. Let's move on and talk about left ventricular assist devices or LVADs. First are those devices that we can use long term, from months to years, either as a bridge to transplant or as destination therapy. There used to be two devices on the market, but the Heartware device was removed. And so now we're left with HeartMate 3. It is a magnetically levitated, centrifugal, continuous flow pump. It generates artificial pulsatility, and it's afterload and preload dependent. That magnetic levitation reduces the amount of friction and reduces the amount of potential thrombus formation. Short term, a device that's used for days to weeks for urgent support as a bridge to therapy or bridge to decision. Typically, we're using the Centromag. Again, a magnetically levitated, centrifugal, continuous pump. It is a disposable pump that can be integrated with an oxygenator and thus used as ECMO. And it provides left ventricular, right ventricular, and biventricular support. This is the HeartMate 3. We see it here, this centrifugal pump that is attached to the apex of the left ventricle. And with that pump, there's this artificial pulse algorithm. So there's a ramp down, a somewhat instantaneous ramp down of the RPMs of the device, which creates a pulse here. We see, again, ramp down and pulse here, which is meant to prevent suck down events. This is what the inflow cannula should look like with the HeartMate 3 device. We see that the inflow cannula is free of the septum and pointed directly towards the mitral valve. This is a 3D view of that. Now, there are three main parameters we want to think about when we're managing a patient with a HeartMate 3 in the ICU. The first is the power. The power is the measure of the pump current and voltage, essentially, it's the work done by the pump to generate flow. The next is flow. And flow is actually a calculated parameter. It uses power and speed, as well as a hematocrit. Finally, we have the pulsatility index, which is a ratio in the difference of the maximal power and the minimal power over the average power. This is a surrogate for intrinsic LV contractility. The pulsatility index is directly related to pump flow. And the pulsatility index will change with changes in preload and afterload. You can get pulsatility index events, or PI events, when there is a change of greater than 45% in the instantaneous pulsatility index versus the average pulsatility index. This slide illustrates how some clinical scenarios that you see in the ICU affect power, flow, and the pulsatility index. Now, importantly, we need to evaluate these changes in power, flow, and pulsatility index in the context of the clinical presentation, labs, and vitals, as well as with imaging, such as echocardiography. In fact, let's look at RV failure, which is typically seen as a decrease in power, a decrease in flow, and variable changes in pulsatility index. So let's see how echocardiography can help us with these low flow alarms, for example, from RV failure. With echocardiography, we should see a decrease in LV size. So that is to say a decrease in the LV in diastolic diameter with an increase in RV size. The interventricular septum and interatrial septum should shift leftward, leftward into the left ventricle. There'll be decreased aortic valve opening. And then you need to see markers of increased right atrial pressure, such as a dilated IVC. Tricuspid regurgitation a lot of times will increase as well due to that dilation of the right ventricle with increased load. So here's a TEE demonstrating the worst case scenario with RV dysfunction. We see that the RV here is dilated and hypokinetic. The interventricular septum has shifted to the left towards the left ventricle, which created a suction event. The left ventricle has sucked down on the impella inflow cannula. We see it here and here. Worst case scenario, with this situation, we need to drop the RPMs and allow the left ventricle to re-expand. This slide demonstrates our ECHO findings in the setting of high flow. Again, thinking about pump thrombosis, peripheral vasodilatation, and severe aortic regurgitation, and the findings that we'll see with ECHO. Let's end our journey through mechanical circulatory support with VA ECMO, which is used as a bridge to recovery, transplant, or to a more durable device. VA ECMO provides comprehensive support for the patient, including oxygenation, ventilation, and cardiopulmonary support. Indications include cardiogenic shock or obstructive shock due to PE, for example, cardiac arrest, respiratory failure, primary graft failure of the heart and or lung, failure to wean from cardiopulmonary bypass, and or as a mechanical support device used as a bridge to recovery, or as a bridge to more advanced therapy, such as a durable VAD. Important contraindications include an irreversible disease process, for example, catastrophic CNS injury, pre-existing severe disease process, for example, hepatic failure or malignancy, or situations that are determined to be futile. Relative contraindications include aortic dissection, or aortic dissection Relative contraindications include aortic dissection or severe aortic regurgitation. Now, the fundamental components of the circuit include the pump, the oxygenator with air blender, and the venous access cannula and return arterial cannula. The venous access cannula is usually 21 to 25 French, and the return cannula is usually 17 to 21 French. The cannulation strategy is really important with VA ECMO. Peripheral cannulation is the most common strategy since cannulas can be placed at the bedside or in the catheterization laboratory. However, the physiology associated with peripheral cannulas increases the risk of some complications. The return cannula delivering continuous flow into the femoral artery increases afterload for the left ventricle. This increased afterload is proportional to the flow, and it can cause reduced pulsatility of the left ventricle. Essentially, the aortic valve can stop opening. The decreased LVE ejection can lead to increased left ventricular distension from blood returning to the left ventricle through the pulmonary and bronchial circulation. The inadequate decompression of the left ventricle then causes increased left ventricular wall stress, which leads to ischemia. The LVEVP, or left ventricular end-diastolic pressure increases, you have an increase in left atrial pressure, and then pulmonary edema, and then thrombus formation can occur in the left ventricle or the aortic root. In this setting, unloading or decompression of the left ventricle is required, and this can be facilitated through the use of an intra-aortic balloon pump, impella device, or a transeptal drainage catheter. Now, the next issue is Harlequin syndrome or differential hypoxia. Essentially, the poorly oxygenated blood from the native circulation mixes with the oxygenated blood from the ECMO circuit in the aorta. In the setting of severe left ventricular dysfunction, the mixing cloud may occur in the ascending aorta. However, as native contractility improves, that mixing cloud may migrate into the descending aorta, which can expose the upper body and head to hypoxic blood coming from the heart. In this situation, the head and the upper body may appear blue, and the lower body will appear pink, thus giving the name Harlequin syndrome. Therefore, with peripheral cannulation, it's important to ensure that the right hand and right side of the brain are being appropriately oxygenated. Therefore, oxygen saturation should be measured on the forehead, ear, and right hand, and ABGs should be assessed from the right radial artery. The other key issue with peripheral cannulation is the presence of a large-bore return cannula in the femoral artery. This increases the risk of limb ischemia. Distal perfusion should be monitored regularly with duplex ultrasound. This is an example of inadequate LV unloading in the setting of VA ECMO with a peripheral cannulation strategy. This patient had a cardiomyopathy and was on VA ECMO's bridge-to-transplant. The patient was having recurrent ventricular tachyarrhythmias and VT storm, and so the ECMO flow was turned up to minimize inotropic and vasoactive medications. In that setting, the patient developed worsening pulmonary vascular congestion with bilateral effusions. We see with ECMO, the LV is quite dilated, and then there is significant mitral regurgitation due to that high left ventricular end-diastolic pressure and the inadequate unloading of the left ventricle. So this patient was taken to the cardiac cath lab and a percutaneous left atrial vent was placed to vent the patient. This slide documents some of the general guidelines for ECMO management. Essentially, ECMO flow should be optimized to normalize distal perfusion. The ECMO circuit requires adequate preload, typically targeting a CVP of at least 10, and the MAP target should be around 65 to 90 in order to reduce excessive afterload on the left ventricle. Targeting a mixed venous oxygen saturation of greater than or equal to 70% with clinical evidence of perfusion. So monitoring for urine output and capillary refill is important. The gas flow or sweep is initially set at one-to-one with the ECMO flow rate. The gas sweep can be titrated based on the pCO2, but we have to bear in mind that we should never reduce the sweep to less than 0.5 liters per minute as this is the minimal amount required for oxygen delivery to the patient. Systemic anticoagulation is required to prevent thrombus formation in the circuit, typically targeting a PTT of 1.5 to two times normal or an ACT of 180 to 220. It's also important to monitor fibrinogen levels. Now, we want to assess for hemolysis and thrombosis, and so daily LDH levels are important. We also need to visually inspect the circuit and oxygenation membrane for evidence of thrombus formation. If bleeding occurs, it's important to guide clotting factor repletion with thromboelastography. Now, there are a lot of different issues that can occur with VAECMO, but perhaps one of the most important ones to be able to recognize and treat is access insufficiency as this can lead to low flows. Essentially, this occurs when the suction pressure at the access cannula is too excessive for venous return and the inflow is interrupted to the pump. And what you can see is shaking or chattering of the circuit tubing. And if you look at the monitor, you'll see very high negative pressures for the venous access cannula. Now, the etiology includes hypovolemia and bleeding. That's probably the most likely cause in most settings, but you also have to be aware of issues with the cannula itself or the presence of cardiac tamponade or tension pneumothorax, which can impair venous return. Acute vasodilation can also do this, as well as increased intra-abdominal pressure and severe aortic regurgitation. So the steps that you want to take, you want to quickly reduce pump speed to alleviate that suction. And then you want to check the cannula. Ideally, you want to use ultrasound at the bedside to be able to assess for a tamponade and to assess for a tension pneumothorax. You can also move to a TEE if imaging is poor. And then you want to volume resuscitate. You want to use your CVP to understand if the patient needs more volume to be able to fill the venous cannula. And then if there's still issues, you may need to place an additional venous cannula. Troubleshooting low flow alarms is also incredibly important. And you need to have a differential to move through at the bedside. Access insufficiency, again, is typically a very common cause. But issues with circuit tubing or oxygenator occlusion may be a problem. Again, bleeding, which leads to access insufficiency, air embolism, or significant elevations in blood pressure. For example, when a patient is waking up off sedation and agitated. In summary, we reviewed the differential diagnosis of hemodynamic instability in the post-cardiac surgical patient. We discussed epicardial pacing wires, the importance of assessing the capture threshold daily as increasing thresholds may be a sign of eventual loss of capture. We need to remember to optimize AV synchrony as much as possible in those patients with epicardial pacing leads as this improves cardiac output. With intra-aortic balloon pumps, we want to utilize WAIM forms to make sure that we are optimizing the hemodynamic benefit of the balloon pump. In those patients with impella devices, we want to use echo regularly to optimize positioning and to troubleshoot any issues. And finally, with VA ECMO, particularly with the peripheral cannulation strategy, we need to remember that this increases LV afterload and these patients may require LV unloading. There's increased risk of differential hypoxia, termed Harlequin syndrome. And then finally, lower extremity ischemia is a big problem. We need to assess for distal pulses daily. And then if there's any sort of risk, we need to push for the placement of a distal perfusion catheter.

post-cardiac surgery

vasoplegia

low cardiac output syndrome

right ventricular failure

cardiac tamponade

mechanical circulatory support

echocardiography

intra-aortic balloon pump

ventricular assist devices