Introduction
Mortality for patients with myocardial infarction (MI) remains unacceptably high despite improvements in medical therapy and revascularization. Death, when it occurs, is often caused by cardiac dysfunction, which leads to low cardiac output, hypotension, and end-organ failure ( Figure 27-1 ). This syndrome is broadly referred to as cardiogenic shock, and can be considered a continuum of severity from pre-shock to refractory shock (see Chapter 25 ). Cardiogenic shock can result from a variety of complications of MI, including left ventricular (LV) dysfunction, mitral regurgitation (MR), ventricular septal rupture, right ventricular failure, or cardiac tamponade (see Chapter 26 ).
Temporary mechanical circulatory support (MCS) devices improve hemodynamic parameters of patients with cardiogenic shock. In theory, timely and appropriate use of MCS will interrupt the shock cascade and prevent multiorgan dysfunction and death (see Figure 27-1 ). Devices such as the intra-aortic balloon pump (IABP), Impella (Abiomed, Danvers, Massachusetts), TandemHeart (Cardiac Assist, Pittsburgh, Pennsylvania), and extra-corporeal membrane oxygenation (ECMO) can be deployed rapidly and can provide clinically significant hemodynamic support. To date, these hemodynamic benefits have not been translated into improved outcomes in randomized trials, and the role of MCS remains uncertain despite decades of clinical use. A recent position paper from the American Heart Association and Society of Cardiac Angiography and Intervention highlights this uncertainty, noting definitive clinical evidence is unavailable or controversial in many cases. Nonetheless, use of MCS is increasing in the United States.
We provide a review of currently available technologies with a detailed look at the advantages and disadvantages of each, a summary of available trial evidence, and guidance for clinical situations in which each device may be helpful. Clinicians should carefully weigh the putative benefit of circulatory support with an increased risk of complications for each patient when persuasive evidence of improved clinical outcomes with MCS is not available.
Mechanical Circulatory Support Devices
The ideal MCS device is able to be placed quickly at the bedside, corrects the hemodynamic deficits resulting from MI, causes few serious complications, and may be removed easily when it is no longer needed. MCS should (1) restore systemic circulation to normalize end-organ function and prevent multisystem shock, (2) increase myocardial blood flow to normal and ischemic coronary territories, and (3) decrease myocardial oxygen demand and limit the extent of ischemia and/or infarct. No current devices fully meet these criteria. Thoughtful use of MCS devices requires an understanding of the hemodynamic effects of each device, the ease of use and frequency of complications, and available clinical trial evidence ( Figure 27-2 ).
Studies of MCS devices have focused on cardiogenic shock caused by LV dysfunction, which is the dominant cause of shock after ST-elevation MI (STEMI) (see Chapter 25 ). Evidence of use for patients with acute MR, right ventricular failure, or ventricular septal defects (VSDs) is explicitly noted, when available.
Intra-Aortic Balloon Pump
Hemodynamic Effects
Aortic counterpulsation is the most widely used and mature technology for hemodynamic support. This support modality uses a balloon positioned in the descending aorta that inflates during diastole and deflates immediately before systole. With inflation, the balloon displaces blood from the descending aorta and increases diastolic pressure. This leads to improved peripheral and coronary artery perfusion. With deflation, early systolic blood pressure falls and results in reduced LV afterload. Overall, mean arterial pressure is increased, mostly because of diastolic augmentation by balloon inflation ( Figure 27-2 ).
The benefits of aortic counterpulsation include (1) increased mean arterial pressure and cardiac output, (2) increased coronary blood flow, and (3) decreased myocardial oxygen demand ( Figure 27-e1 ). As such, IABP fits the hemodynamic criteria for an ideal MCS device. However, the impact of IABP on cardiac output is generally small and is dependent on the loading conditions and function of the LV. No more than a 0.5 to 1.0 L/min increase in cardiac output should be expected. Similarly, improvement of coronary perfusion is likely modest and dependent on patient factors. IABP delivers an increase in blood flow to the aortic root (6.4% of balloon volume on 1:1 assistance) and improves flow in open coronary arteries, but the effect on blood flow past coronary stenosis or in patients with acute coronary syndromes is unclear. IABP improves coronary blood flow after thrombolysis and can increase the rate of clot lysis in canine models. Independent of coronary blood flow effects, IABP reduces myocardial ischemia by decreasing oxygen demand. Decreased demand results from a decrease in LV afterload and wall stress, and may be the dominant hemodynamic effect of IABP in the setting of active ischemia.
Clinical Use
IABP is commonly placed percutaneously via the femoral artery, although axillary artery or subclavian artery access is also possible using a surgical cutdown. The IABP is placed over a guidewire and positioned in the descending aorta just distal to the origin of the left subclavian artery (2 to 4 cm below the aortic arch). Modern devices provide automatic balloon inflation based on electrocardiographic or hemodynamic triggers. Timing can be adjusted manually. Inflation should occur at the dicrotic notch and fully deflate before the onset of systole. Inappropriate device timing will limit hemodynamic benefits of the device ( Figure 27-3 ). One inflation per cardiac cycle (1:1) provides maximum hemodynamic support. A reduced assist ratio (1:2 or 1:3) may be more appropriate for patients with tachycardia, arrhythmia, or for weaning before device removal. Systemic anticoagulation is commonly used when using IABP with an assist ratio of less than 1:1. However, systematic evaluations of the effects of anticoagulation versus placebo are lacking. Device use at 1:1 for limited periods (<24 hours) without anticoagulation is probably safe and is often used to allow anticoagulation to dissipate before IABP removal. Moderate or severe aortic insufficiency may worsen with IABP use and is an absolute contraindication, as is the presence of aortic dissection. Relative contraindications include severe peripheral arterial disease, aortic aneurysm, bleeding, or inability to safely administer a systemic anticoagulant. The IABP exists in several sizes, from 34 to 50 mL. The most commonly used size is 40 mL. Recently, the 50-mL IABP has been evaluated, and it may be associated with slightly better hemodynamic augmentation compared with the 40-mL device.
An IABP is relatively easy to insert and has a low complication rate. The Benchmark Counterpulsation Outcomes Registry prospectively registered 5495 patients who received an IABP in the setting of MI from 1996 to 2001. Balloon insertion was successful in 97.7% of patients, and in-hospital mortality was 20%. Only 2.7% of patients experienced major complications (severe bleeding [1.4%], major limb ischemia [0.5%], balloon leak [0.8%], or death related to IABP [0.05%]). Prolonged use of IABP is also associated with thrombocytopenia, hemolysis, and infection.
In real-world practice, MI complications account for only a minority of IABP insertions. In the Benchmark Registry, 27.3% of patients received an IABP because of cardiogenic shock, and 11.7% received it for mechanical complications of MI. Although use of MCS is increasing overall in the United States, there has been a slight decline in IABP use since 2008.
Observational Evidence
Kantrowitz and Kantrowitz suggested the concept of balloon counterpulsation in 1953, and the first clinical use occurred in 196 7 . The IABP gained popularity in subsequent decades, supported by promising observational evidence. The GUSTO-I trial included 2972 patients with STEMI complicated by cardiogenic shock. Of these, the 25% of patients treated with early IABP demonstrated a trend toward lower 30-day mortality compared with patients not treated with IABP or treated later in the clinical course (47% vs. 60%; P = .06). Use of an IABP was 86% in both arms of the SHOCK trial, which showed benefit of early revascularization compared with medical management for patients with STEMI and cardiogenic shock. In the SHOCK trial registry, 51% of patients were treated with IABP, and IABP use was associated with lower mortality (50% vs. 72%; P < .0001). Among 23,180 patients with acute MI and cardiogenic shock in the NRMI-2 cohort, 31% were treated with IABP. IABP use was associated with reduced mortality in patients who received thrombolytic therapy, but who did not undergo primary angioplasty.
In 2004, IABP use for patients with MI complicated by shock became a class I recommendation despite a lack of randomized data. A later meta-analysis of observational data showed no association of IABP therapy with survival for patients treated with percutaneous coronary intervention (PCI) in the setting of STEMI with cardiogenic shock, although IABP was associated with survival among patients treated with thrombolysis.
Randomized Trials
Subsequently, several randomized trials have addressed this question ( Table 27-1 ). The IABP SHOCK I trial randomized patients with MI and shock to IABP or no IABP. This small trial assessed hemodynamic endpoints only. Temporal improvements in cardiac output and systemic vascular resistance were seen in patients managed both with and without IABP, with no significant difference between groups, which is in contrast to other physiological studies. The IABP-SHOCK II study then randomized 600 patients with cardiogenic shock that complicated STEMI to IABP or medical therapy at the time of PCI. There was no significant difference between groups in any clinical endpoints at 30 days or 1 year. These neutral results were consistent across all of the major subgroups in which a benefit of IABP would have been plausibly more likely. The CRISP-AMI trial randomized 337 patients with anterior STEMI to elective versus provisional insertion of IABP. Although performed in patients with stable hemodynamics, this study tested the hypothesis that IABP might reduce the area of MI in patients at risk of cardiogenic shock. The mean infarct size, as assessed by cardiac magnetic resonance imaging, was not significantly different between groups. Clinical endpoints were similar at hospital discharge and 30 days. A nonsignificant trend toward a mortality benefit with IABP was seen at 6 months.
| Study | Period | Sample Size | Setting | Revascularization Type | Hemodynamic Outcomes | Clinical Outcomes |
|---|---|---|---|---|---|---|
| IABP vs. Medical Therapy | ||||||
| Ohman, 2005 (TACTICS) | 1996–1999 | 57 | Multicenter | Thrombolysis (100%), PCI (23%), CABG (18%) | Not reported | Mortality at 6 mos similar for IABP and control (34% vs. 43%; P = .23) Trend toward benefit for IABP among patients with Killip III/IV classification |
| Prondzinsky, 2010 (IABP SHOCK I) | 2003–2004 | 40 | Single-center | PCI (100%) | No significant difference in cardiac output, SVR, PCWP | No difference in APACHE II score |
| Thiele, 2012 (IABP SHOCK II) | 2009–2012 | 598 | Multicenter | PCI (96%), CABG (4%), None (3%) | No difference in HR, BP, or serum lactate | Mortality at 30 days similar for IABP and control (39.7% vs. 41.3%) |
| TandemHeart vs. IABP | ||||||
| Thiele, 2005 | 2000–2003 | 41 | Single-center | PCI (95%), CABG (5%) | Patients treated with TandemHeart had greater improvements in cardiac power index, PCWP, CVP, and serum lactate | Mortality at 30 days was similar for patients treated with TandemHeart and IABP (43% vs. 45%; P = .86) |
| Burkhoff, 2006 | 2002–2004 | 33 | Multicenter | Among AMI patients: PCI (85%), CABG (12%) | Patients treated with TandemHeart had greater improvements in CI, MAP, and PCWP | No difference in 30-day mortality. |
| Impella vs. IABP | ||||||
| Seyfarth, 2008 | 2004–2007 | 26 | Two centers | PCI (94%), CABG (4%) | Patients treated with Impella had significantly improved CI | No difference in 30-day mortality (46% in both arms) |
In summary, current observational evidence supports use of IABP for patients with post-MI shock who are receiving thrombolytics. Routine use of IABP for MI patients who are undergoing PCI is not supported by current observational or randomized trial data ( Table 27-2 ; see also the section on Suggested Approach ). A provisional strategy of selective use for refractory shock has not been tested. Use of IABP has been described for patients with acute VSDs and MR with cardiogenic shock (see Chapter 26 ). In one series, IABP support was associated with lower preoperative mortality. IABP may have a role in stabilization of shock to permit definitive surgical correction.
| German-Austrian S3: Cardiogenic Shock (2012) | ACCF/AHA: STEMI (2013) | ESC/EACTS: Revascularization (2014) | |
|---|---|---|---|
| IABP: Shock | In the setting of lytic therapy: IABP should be carried out adjunctively. In the setting of PCI: May be considered, but the available evidence is unclear. | IIa, B IABP can be useful for patients with cardiogenic shock after STEMI who do not quickly stabilize with pharmacological therapy. | III, A Routine use of IABP in patients with cardiogenic shock is not recommended. |
| IABP: Mechanical complications of MI | No recommendation | No formal recommendation. “IABP can provide temporary circulatory support.” | IIa, C IABP insertion should be considered in patients with hemodynamic instability due to mechanical complications. |
| TandemHeart, Impella: Shock | No recommendation | IIb, C Alternative LV assist devices for circulatory support may be considered in patients with refractory cardiogenic shock. | IIb, C Short-term mechanical circulatory support in ACS patients with cardiogenic shock may be considered. |
TandemHeart
Hemodynamic Effects
TandemHeart is a percutaneous ventricular assist device (VAD) that can provide up to 4 L/min of circulatory support, making it attractive for patients with poor intrinsic systolic function and/or prolonged anticipated periods of shock. TandemHeart draws blood from the left atrium and injects it into the iliac artery or abdominal aorta. It consists of a left atrial drainage catheter, an extracorporeal centrifugal pump, and a femoral artery inflow catheter ( Figure 27-2 ).
For patients with cardiogenic shock, TandemHeart results in a reduction in pulmonary capillary wedge pressure (PCWP) and pulmonary artery pressure, augmented systolic blood pressure, and an improved cardiac index ( Figure 27-4 ; also see Figure 27-2 ). Compared with IABP, TandemHeart provides superior left heart preload reduction because of direct aspiration of blood from the left atrium, significantly reducing LV volume and pressure (see Figure 27-e1 ), which results in greater reduction in the PCWP. The cardiac index was also improved more effectively with TandemHeart. The effect of TandemHeart on coronary perfusion is unknown.
Clinical Use
TandemHeart was first described in 2001 and received US Food and Drug Administration (FDA) 510(k) approval in 2006. Use is less common than IABP because of its greater complexity of deployment and a higher complication rate. A 21F venous catheter is advanced via the right femoral vein to the right atrium ( 
). A transseptal puncture under fluoroscopic and/or echocardiographic guidance introduces this inflow cannula into the left atrium. A 15F to 17F catheter is placed via percutaneous puncture of the common femoral artery and positioned in the iliac artery or distal aorta. The extracorporeal pump then provides up to 4 L/min of cardiac output at speeds of up to 7500 rpm. Bilateral 12F femoral artery cannulae may be used instead of the 17F cannula, but this will limit overall flow to 3 L/min. Device placement, even in experienced hands, can take between 15 minutes and 1 hour. Complications are common. In a series of 117 patients with refractory cardiogenic shock, the most frequent complications included sepsis (29.9%), bleeding around the cannula site (29.1%), gastrointestinal bleeding (19.7%), coagulopathy (11.0%), stroke (6.8%), groin hematoma (5.1%), and limb ischemia (3.4%). Overall, 59.8% of patients received blood transfusions. In a population of patients who received short-term support for elective high-risk PCI, major vascular complications occurred in 13% of patients, and thrombocytopenia in 10%. In addition, the inflow cannula can migrate back into the right atrium, which requires repositioning to avoid delivery of unoxygenated blood from the right atrium to the systemic circulation. Cardiac tamponade and perforation can also occur. Anticoagulation is mandatory, with a target partial thromboplastin time of 60 to 80 seconds.
Clinical Trial Evidence
Two randomized studies have compared TandemHeart to IABP in patients with cardiogenic shock ( Table 27-1 ). Thiele and colleagues randomized 41 patients with shock following MI to either TandemHeart or IABP. The TandemHeart group demonstrated greater improvement in the cardiac power index, cardiac output, PCWP, pulmonary artery pressure, and serum lactate. There was no significant difference in mortality, and the TandemHeart group had a higher incidence of limb ischemia (33% vs. 0%), bleeding requiring transfusion (91% vs. 40%), and disseminated intravascular coagulation (62% vs. 15%). In a similar study by Burkhoff and colleagues, 33 patients with shock, 70% presenting with MI, were randomized to TandemHeart or IABP. Again, patients randomized to TandemHeart had a significantly improved cardiac index and PCWP, with no change in clinical outcomes. Adverse events were balanced between the two groups. There are no studies that have compared TandemHeart to medical therapy or MCS other than IABP. Evidence of successful treatment of patients with mechanical complications is lacking. Use of TandemHeart for right ventricular support via right atrium and pulmonary artery cannulation has been described.
In summary, this device provides a high level of cardiac output, but with high complication rates. TandemHeart might be considered in centers where there is substantial experience with its use.
Impella
Hemodynamic Effects
The Impella devices are microaxial flow rotary pumps that are deployed across the aortic valve, drawing blood from the LV and depositing it in the ascending aorta. The Impella device is available in several sizes for support of the left heart (2.5, CP, LD, and 5.0) and the right heart (RP). The Impella 2.5 provides up to 2.5 L/min of augmented cardiac output. In the limited published data with the larger device, the Impella CP, cardiac output augmentation of up to 3.5 L/min has been reported. The Impella 5.0 and Impella LD (placed directly into the aorta) provide up to 5 L/min of support, but require surgical placement. Like other MCS devices, the Impella devices increase cardiac output and mean arterial pressure and decrease PCWP ( Figure 27-5 ). The Impella unloads the LV directly, which results in an immediate reduction in end-diastolic wall stress (see Figure 27-e1 ). Coronary perfusion pressure is increased, potentially because of an elevation of aortic pressure and decreased LV intramyocardial pressure. Unlike the IABP, augmentation of cardiac output is independent of native cardiac function, making it a useful device for patients with moderate to severe reduction in cardiac performance. In the USpella registry of 154 patients with MI and cardiogenic shock across 38 US hospitals, hemodynamic parameters improved after deployment of Impella, including mean arterial pressure (94 mm Hg vs. 63 mm Hg), PCWP (19 mm Hg vs. 32 mm Hg), and cardiac index (2.7 L/min/m 2 vs. 1.9 L/min/m 2 ). The Impella RP is placed via femoral venous access and provides up to 5 L/min of support.
