Thank you for staying until the end. The last talk is on mechanical circulatory support. As before, I have no disclosures. Here's a view of the spectrum of available devices. Of the surgical ventricular assist devices, the HeartMate 3 is in most common use, but we won't be considering durable VADs in detail in this talk. We will instead focus on percutaneous mechanical circulatory support. The next generation promises to be smaller and better, but that's a story for another day or perhaps another year. This slide shows trends in short-term mechanical circulatory support up to 2011. Intra-air balloon pumps and Pluit at the top were still in widest use, but there was a sharp uptick in short-term percutaneous devices. This more recent study, separated out by age group, shows a decline in the use of IABP, but a continuation of the sharp rise in use of MCS devices, including Impella Tandem Heart on the left and ECMO on the right. There are a number of potential indications for mechanical circulatory support. Cardiogenic shock post-myocardial infarction, post-cardotomy, decompensated chronic heart failure, in which case transplant eligibility impacts both the strategy and choice of device, myocarditis, and ventricular arrhythmias. Types of hemodynamic support are listed here by indication. Temporary support may provide a bridge to recovery, or one device may be used to stabilize a patient until they can be bridged to another device, so-called bridge-to-bridge. The devices may be used as triage, as a bridge to a decision about whether further mechanical support or more advanced mechanical support is warranted. Mechanical support can be used to support organ function as a bridge to transplantation, or it can be implanted as permanent support, so-called destination therapy. I will be talking about effects of various mechanical circulatory support devices on various hemodynamic indices that are shown on this slide of pressure-volume loops, which you have seen earlier today. The area inside the pressure-volume loop represents stroke work. Left ventricular wall stress is represented by systolic blood pressure multiplied by end-diastolic volume. The end-systolic point of pressure-volume loops under different loading conditions forms a line whose slope is an index of contractility. The slope is a pressure divided by a volume, and so is an elastance. Volume over pressure is compliance, it's reciprocal, pressure over volume is elastance. The change in pressure divided by the change in volume, or arterial elastance, is proportional to afterload. Finally, myocardial oxygen consumption is measured by stroke work added to potential energy, which is the area inside this triangle. Let's use pressure-volume loops to describe the physiology of heart failure and cardiogenic shock. On the left is a normal pressure-volume loop with normal contractility and afterload. With heart failure, the contractility is reduced and the pressure-volume loop is shifted to the right. The afterload goes up a bit to maintain blood pressure, and the stroke volume is somewhat decreased with an increase in left ventricular volume and end-diastolic pressure. In cardiogenic shock, contractility is further reduced, afterload goes up but is no longer able to maintain systolic pressure, and stroke volume is markedly reduced. The various percutaneous mechanical circulatory support devices are shown here. The intracorporeal pumps include the intraortic balloon pump, which is pulsatile, and the axial flow pump, which is continuous. The two extracorporeal flow pumps are centrifugal. We'll start by talking about the intraortic balloon pump. The IABP is implanted just distal to the subclavian artery. It inflates at the start of diastole, right at the dicrotic notch, augmenting diastolic coronary perfusion pressure, and deflates in systole, reducing afterload and aortic end-diastolic pressure. As you have just seen, the intraortic balloon pump is a volume displacement pump. Left ventricular pressure decreases and stroke volume increases a little bit. On average, the intraortic balloon pump increases cardiac output by a third, while the diastolic balloon pump increases cardiac output by about half a liter. The effects of the intraortic balloon pump include decreased left ventricular pressure, decreased end-arterial elastance, decreased end-diastolic pressure, decreased stroke work, and decreased myocardial oxygen consumption. The IABP does not change contractility, and systolic elastance is unchanged. The tandem heart device is inserted through the femoral vein across the foramina valley via transeptal puncture, and powered by an extracorporeal centrifugal pump. It has continuous flow and removes oxygenated blood from the left atrium, and returns it through the femoral artery into the arterial circulation. The continuous axial pump is inserted into the left ventricle across the aortic valve. There are four versions, with different sizes and maximal flows. Inflow from the catheter in the left ventricle reduces left ventricular end-diastolic volume and end-diastolic pressure, thus reducing myocardial oxygen demand and improving supply-demand balance. Outflow into the aorta from the pump increases aortic pressure and myocardial oxygen supply, and also increases cardiac output. The effects of percutaneous left ventricular assist devices, either centrifugal or axial, on pressure-volume loops are shown on this slide. They provide anywhere between 2.5 and 5.5 liters per minute of cardiac output. They decrease afterload, as shown by the shift in the end-arterial elastance curve, and also decrease myocardial oxygen consumption, as shown by the shift in the end-arterial elastance curve. This is shown by the movement of the pressure-volume loop downward and to the left. Extracorporeal membrane oxygenation is a continuous flow device that removes blood from the right atrium and oxygenates it using a membrane oxygenator. The oxygenated blood is returned to the descending aorta through the femoral artery. In some patients whose left ventricular function is so diminished that the aortic valve doesn't open, an injection of the left ventricle may be necessary due to continued blood return from the bronchial circulation. ECMO can provide up to 7 liters per minute of cardiac output, a value greater than that of other percutaneous devices. This pressure-volume loop shows that the transfer of blood from venous to arterial circulation increases systolic pressure, which may be good for coronary perfusion, but also increases left ventricular afterload, which may not be so good for left ventricular recovery. This slide shows the levels of support offered by different devices, from intra-aortic balloon pump on the low end to VA ECMO on the high end, with the impella and the tandem heart somewhere in between. Here are some of the considerations of mechanical circulatory support devices. Aortic counterpulsation with an intra-aortic balloon pump is easy to use and has low costs compared to the other devices, and it's fairly widely available. However, there is a limited increase in cardiac output as we've mentioned before, and it does not unload the left ventricle directly. The left atrium to aorta circuit, the tandem heart, provides left ventricular unloading and a high degree of support. However, it increases the afterload. There are technical challenges with transeptal puncture and occasionally challenges with stability of the left atrial cannula. ECMO is a right atrium to aortic circuit, also provides a high degree of support and oxygenates the blood as well as circulates it. However, it does not unload the left ventricle and in fact increases afterload, and there is the potential for north-south syndrome with recovery of left ventricular function where oxygenated blood goes to the lower extremities, but the native heart pumps relatively deoxygenated blood to the brain. The left ventricle to aorta axillary pumps, the impella, provides direct left ventricular unloading and graded left ventricular support, but they're expensive. The rapid pump can cause hemolysis, and there are access issues with downstream of the femoral cannulation at times. This is a schematic of the pathophysiology of cardiogenic shock. Myocardial dysfunction predisposes to ischemia, which can cause progressive myocardial dysfunction and lead to death if not reversed. Systolic dysfunction decreases cardiac output and stroke volume, causing hypotension, which can compromise coronary perfusion pressure, and decreased systemic perfusion can cause compensatory vasoconstriction, increasing afterload, and fluid retention, increasing preload, and causing pulmonary congestion. That pulmonary congestion can cause hypoxemia and again predispose to ischemia. Now, mechanical circulatory support can address the pathophysiology of cardiogenic shock. These are data from randomized trials put together by Hoga Tila, and you can see that mechanical circulatory support can improve cardiac index, it decreases pulmonary capillary wedge pressure, and improves mean arterial pressure. It also improves systemic perfusion as shown by decreases in arterial lactate. Randomized controlled trials, however, have not been shown to improve mortality. On the top are data from the meta-analysis I just showed you of hemodynamic improvements in the randomized trials of mechanical support. And what you can see is that in contrast to the hemodynamic improvement, these trials show no difference in mortality. On the bottom is a comparison of registry data from the mechanical circulatory support trials in red compared to matched patients from the IABP shock trial. So these patients are propensity-matched to have at least approximately equivalent risks, and what you can see is that the results are no different. Here's another analysis of matched pairs from the NCDR database from 2015 to 2017. And what you can see is that, if anything, this analysis favors the intra-aortic balloon pump with respect to not only mortality but major bleeding, and it didn't seem to depend on whether the device was placed before PCI or after PCI. Now, I'm not sure these data suggest that the intra-aortic balloon pump is, in fact, better than a microaxial assist device, but they certainly don't support the contention that the assist device is superior. This is a graph of the device selection algorithm. First, you maximize medical therapy. If hemodynamics improve, you assess salvage potential and whether there are contraindications to devices. Then the device choice depends on the hemodynamic indication. Isolated right ventricular failure, biventricular failure, left ventricular failure, and the presence of coexisting acute lung injury or severe ARDS. Finally, you evaluate for resolution of end-organ damage and then for improvement in cardiac function. There are a number of potential complications of mechanical support devices, including infection, thromboembolism and bleeding, thus requiring a balance between too much anticoagulation and not enough, device failure, and multiple organ system failure. Patient selection is the key to interpretation of results of trials of mechanical circulatory support. Survival in acute coronary syndrome with cardiogenic shock with medical therapy is about 40 to 50%. If mechanical circulatory support is applied, you wind up with three possible groups, one of which is the patient, and the other is the organ. If mechanical circulatory support is applied, you wind up with three possible groups, three possible cohorts, A, B, and C. Cohort A are patients who would have survived regardless, and cohort C are patients who would have succumbed regardless, and so the results of the trial, the device can only make a difference in cohort B, and what you have are patients who survived but you also have patients who would have survived, but in fact, succumbed because of mechanical circulatory support. You can't really distinguish these two, but the results of the trial depend on more patients who survive because of your therapy than succumb because of your therapy. Now, if you, in a registry or in a trial, if you add patients who survive regardless, you will make your therapy look better than it really is, and if you add patients who will succumb regardless, you will make patients look worse than it really is. As such, in order to really get an interpretable result of the real benefits of mechanical circulatory support in a population, you need a randomized controlled trial. There is such a trial underway. The Danish shock trial added the Germans and became the Danger Shock Trial, the nice name of the trial. 360 patients with MI and cardiogenic shock are being randomized to open-label impella or guideline-driven therapy with a protocol for escalation, as you can see on the right. As of last count, there were slightly north of 210 patients randomized so far. The COVID epidemic decreased randomization numbers, but they are picking up now. Patients comatose after out-of-hospital cardiac arrest are excluded, and the primary endpoint is all-cause mortality at six months, and I think we're all looking forward to results of this trial. So, to conclude, devices are getting smaller, simpler, and more reliable. Patient selection and timing are the key to success. You want to match the device to the indication and to the physiology and pathophysiology. And it's always important to make sure you have a plan about where you are intending to go.

mechanical circulatory support devices

percutaneous devices

hemodynamic indices

patient selection

randomized controlled trials