Share this article on:

Which Is Better: A Miniaturized Percutaneous Ventricular Assist Device or Extracorporeal Membrane Oxygenation for Patients With Cardiogenic Shock?

Chamogeorgakis, Themistokles*†‡; Rafael, Aldo*; Shafii, Alexis E.*†; Nagpal, Dave*; Pokersnik, Julie A.*; Gonzalez-Stawinski, Gonzalo V.

doi: 10.1097/MAT.0b013e3182a8baf7
Clinical Cardiovascular/Cardiopulmonary Bypass

The purpose of this study is to compare outcomes associated with the use of Impella and TandemHeart short-term support devices with venoarterial extracorporeal membrane oxygenation (ECMO) therapy for postinfarction- or decompensated cardiomyopathy–related cardiogenic shock. Between January 2006 and September 2011, 79 patients were supported with either an Impella axial flow pump (n = 7) or a TandemHeart centrifugal pump (n = 11), or with ECMO (n = 61) therapy for cardiogenic shock in a single institution. Pertinent variables and postprocedural events were analyzed in this cohort of patients using a prospectively maintained clinical database. The in-hospital mortality, successful weaning from mechanical circulatory support, bridge to long-term destination support device and heart transplantation, and limb complications did not differ between the 2 groups based on intention-to-treat analysis. Age was the only independent predictor for in-hospital survival. In this cohort of patients, short-term support devices and ECMO achieved comparable results. In the modern era of medical cost restraints, ECMO may be more cost effective for patients with postinfarction- or decompensated cardiomyopathy–related cardiogenic shock. Larger randomized trials may be necessary to further elucidate this topic.

From the *Department of Thoracic and Cardiovascular Surgery, Cleveland Clinic Foundation, Cleveland, OH; Department of Cardiothoracic Surgery, Baylor University Medical Center, Dallas, TX; and National and Kapodistrian University of Athens, Athens, Greece.

Submitted for consideration March 2013; accepted for publication in revised form August 2013.

Disclosure: The authors have no conflicts of interest to report.

Reprint Requests: Gonzalo V. Gonzalez-Stawinski, MD, Cardiac Surgery, Baylor University Medical Center, 3900 Junius Street, Suite 605, Dallas, TX 75246. Email:

The field of cardiocirculatory support has undergone a considerable evolution since the first heart–lung machine was created by Gibbon in 1960.1,2 At the time, Gibbon’s invention was revolutionary and propelled the field of cardiac surgery. Although Gibbon’s creation allowed for the care of the cardiac patient in the operating room, its applicability outside the operating theater was largely unimaginable at the time. Eventually, the need to provide mechanical circulatory support (MCS) expanded outside the operating room, resulting in the development of extracorporeal membrane oxygenator (ECMO) technology. However, innovations in the MCS field progressed, leading to the development of smaller implantable pumps compact enough to be inserted through a peripheral vessel but powerful enough to maintain end-organ perfusion. Two examples of these miniaturized devices are the TandemHeart and the Impella. Each of these has shown to be safe and effective in staving off the detrimental effects of cardiogenic shock in a variety of situations.3,4

Although these smaller devices have been shown to be better at providing circulatory support compared with less powerful modes of MCS (i.e., intra-aortic balloon pumps), evidence of their performance compared with that of ECMO is scant.5 Therefore, the objective of this study was to determine the outcomes associated to the use of a temporary miniaturized percutaneous ventricular assist device (mp-VAD) compared with ECMO therapy in patients with cardiogenic shock.

Back to Top | Article Outline

Materials and Methods

A retrospective chart review was conducted on all patients who received either ECMO or a short-term mp-VAD between January 2006 and September 2011 at the Cleveland Clinic Foundation in Cleveland, OH. All patients in this study had either postinfarction or decompensated cardiomyopathy (ischemic or nonischemic) cardiogenic shock. In this analysis, the mp-VAD was a TandemHeart (Cardiac Assist, Inc., Pittsburgh, PA) or an Impella 5 LP (ABIOMED, Danvers, MA). Charts provided demographics, indication for support, duration of support, and limb complications related to device placement. Data from charts were additionally used to determine stroke, renal failure, weaning rates, rates of bridging to a long-term mechanical assist device or heart transplantation, and in-hospital as well as mid-term survival. One patient who was initially supported with TandemHeart crossed over to the ECMO group because of inadequate end-organ perfusion; this patient was analyzed based on the initial intention to treat. This study was conducted after the written approval of the Institutional Review Board.

Back to Top | Article Outline

Technical Considerations and Patient Management

ECMO is established using peripheral cannulation with a 20–24 F venous femoral cannula for drainage and axillary artery with a side graft or the femoral artery with direct cannulation of the vessel and distal perfusion of the ipsilateral limb. TandemHeart is placed percutaneously through the femoral vessels for left ventricular support.6 After a transseptal puncture, a venous inflow cannula is inserted in the left atrium. Oxygenated blood is drawn from and returned via a centrifugal pump in the femoral artery. In right ventricular support with TandemHeart, venous drainage is achieved with a 20–24 F venous cannula and blood return in the pulmonary artery with a long percutaneous cannula. The Impella device consists of an axial flow pump; the inflow tip is positioned into the left ventricle and the outflow in the aorta.5 The Impella 5 LP is inserted through the femoral artery with an open exposure of the vessel. A heparin drip is initiated at 400 units per hour targeting for thromboplastin time of 55–65 seconds. For patients receiving chest compressions, ECMO support is the only option. For more stable patients, mp-VAD placement under fluoroscopic guidance is another option. For isolated right ventricular failure, we favor TandemHeart support. In left ventricular failure, Impella 5 LP or TandemHeart can be used.

Patients are admitted to a MCS and transplant intensive care unit. Nurses manage them based on protocols. Perfusionists are available, should the oxygenator need to be replaced in patients on ECMO therapy. In addition, surgical and cardiology heart failure teams round daily on them, formulating a treatment plan. When appropriate hemodynamic function on minimal inotropic support is maintained, separation from MCS is considered. If weaning is not possible, the patient is prepared for long-term device support or heart transplantation. In some patients, it is necessary to withdraw support with the knowledge that support is futile and survival unlikely.

Back to Top | Article Outline


The definition of cardiogenic shock is based on clinical and hemodynamic criteria. The clinical criteria are hypotension (systolic blood pressure <90 mm Hg) and end-organ hypoperfusion (cool extremities and oliguria). The hemodynamic criteria are a cardiac index of <1.8 L/min/m2 and a pulmonary capillary wedge pressure of >20 mm Hg. For postinfarction shock patients, myocardial infarction is documented during the same hospital admission. Patients with a previously established diagnosis of cardiomyopathy (ischemic and nonischemic), clinical deterioration leads to cardiogenic shock. Hyperperfusion syndrome is defined as an ipsilateral, to the side of arterial cannulation, edematous limb that is hyperemic and warm to touch. Compartment syndrome is the more serious presentation of hyperperfusion syndrome in which high compartment pressure causes secondary extremity ischemia. Extremity ischemia is defined as cold, pale, and or mottled ipsilateral extremity due to decreased arterial blood flow. Stroke is defined as any central neurologic deficit diagnosed by clinical exam and/or by radiologic imaging. Finally, renal failure is determined when serum creatinine is >2 g/dL.

For the purposes of this analysis, bleeding is considered if exploration of the artery cannulation site is necessary.

Back to Top | Article Outline


Continuous variables were presented with mean ± standard deviation and compared with student’s t-test or one-way analysis of variance where appropriate. Discrete variables were summarized by percentages and compared with the χ2 test for trend. Multivariate regression analysis for in-hospital survival for the entire group of patients who received mechanical circulatory support was undertaken. The variables of the entire group of patients were evaluated first using univariate analysis. Long-term survival was assessed using the Kaplan–Meier method, and survival curves were compared with the log-rank test, with adjustment for trend. For purposes of this analysis, these patients were analyzed based on the initial intention to treat. All analyses were performed using the SPSS-15 (SPSS, Chicago, IL). All p values were considered significant if <0.05 and were two-tailed.

Back to Top | Article Outline


Between January 2006 and September 2011, 79 patients underwent short-term MCS for cardiogenic shock associated with either postinfarction (n = 46, 58%) or decompensated cardiomyopathy (ischemic, n = 8, 10%; nonischemic, n = 25, 32%). Patient demographics are presented in Table 1. A total of 61 (77%) patients underwent ECMO support, whereas 18 patients (23%) were supported with an mp-VAD device (Impella 5 LP, n = 7; TandemHeart, n = 11). Fifteen patients had Impella (n = 7) or TandemHeart (n = 8) for left ventricular support. Three patients received TandemHeart for isolated postinfarction right ventricular failure.

The mean length of support for the entire group was 4.5 days (range, 1–19 days). Twelve patients (15.2%) developed limb complications related to the arterial cannulation site. Six patients developed limb ischemia, 2 had bleeding at the arterial cannulation site, 2 developed extremity compartment syndrome requiring fasciotomy, and another 2 had hyperperfusion syndrome of the limb that was treated conservatively with limb elevation. The incidence of limb complications was not different between the 2 groups (mp-VAD, n = 4 vs. ECMO, n = 8; p = 0.454; Table 2).

A total of 18 patients (15.2%) were successfully weaned from MCS (ECMO, n = 12 vs. mp-VAD, n = 6; p = 0.336), and of those, 13 (16.5%) survived to hospital discharge (ECMO, n = 9 vs. mp-VAD, n = 4; p > 0.999). Thirty-one patients (39.2%) were successfully bridged to either ventricular assist device or total artificial heart (ECMO, n = 19 vs. mp-VAD, n = 5; p > 0.999) or heart transplant (ECMO, n = 7 vs. mp-VAD, n = 0; p = 0.341), and of those, 26 (33%) survived to hospital discharge. The overall in-hospital survival was 49.4% (n = 39; ECMO, n = 30 vs. mp-VAD, n = 9; p > 0.999).

There was no difference between the two groups with regard to in-hospital mortality (p > 0.999), success to weaning from MCS (p = 0.336), or bridge to long-term device or transplantation (p = 0.288; Table 2).

In the univariate regression analysis, age, success to weaning from MCS, and bridge to long-term support or transplant were the only independent predictors for in-hospital survival. In the multivariate analysis, bridge to long-term support or transplant was the only independent predictor for in-hospital outcome (Table 3). The mean follow-up of the entire cohort was 14.3 months (range, 0–20.4 months). The long-term survival analysis did not reveal any difference between the two treatment groups (Figure 1).

Back to Top | Article Outline


Data presented in this report elucidate the survival benefit associated with using either ECMO or mp-VAD in managing patients with cardiogenic shock and would suggest that the use of either technology may be rationalized simply on operator preference. In our experience, there was no difference with regard to survival with either device; therefore, it is important to understand the merits and flaws of each of these so that an individualized approach can be used when practitioners are faced with certain patient groups.

The merits of the TandemHeart are that it is a small pump, which is easy to prime, and it can be inserted percutaneously. The device is designed mainly for left-sided support, but right-sided configuration can be achieved in certain instances, providing the opportunity for use as a bi–mp-VAD.7 Patients on this device need to be anticoagulated for proper pump function. The major limitations associated with this device include bleeding from the cannulation site, hemolysis, and limb ischemia.4 Additionally, there is the need for a septotomy for its application as a left-sided device, which may limit its applicability to centers with expertise in that procedure. Also, the need for radiologic imaging for its implantation constitutes another limitation.

Although the TandemHeart can be applied to any patient with any form of left ventricular failure, we believe it is better reserved for patients with restrictive cardiomyopathies, and for cases of complications of acute myocardial infarction (i.e., postinfarct ventricular septal defects). The latter example is an important one as the use of an Impella in that situation may be dangerous due to the fragility of the left ventricular wall following an infarct and the potential of ventricular perforation. Another attractive quality of the TandemHeart that the Impella system lacks is that, in cases in which respiratory failure ensues despite adequate pump function, the system can be configured to an ECMO by simply positioning the drainage cannula in the right atrium and inserting an oxygenator into the system.

The Impella is another miniaturized VAD that can be percutaneously implanted. Although two different designs are available, the more effective of the two for cases of cardiogenic decompensation is the Impella 5.0 LP. The Impella 2.5 LP is often unable to sustain adequate end-organ perfusion. Unlike the TandemHeart, the Impella can be used at this time only for univentricular support. However, when compared with ECMO, the Impella is potentially better at ventricular unloading, as reported by Kawashima et al.8 As with the TandemHeart, the major limitations associated with the Impella include bleeding from the access site, hemolysis, and limb ischemia.9 Additionally, to implant this pump, radiologic capability is needed. Because the Impella can be used only as a left-sided device, it is likely better suited for cases of dilated ischemic or nonischemic cardiomyopathies. In these cases, proper pump function may be facilitated by the size of the ventricle.

A merit of ECMO is its ease of implantation. The device can be inserted without radiologic assistance with the Seldinger technique in cases in which biventricular support is needed. As with all other devices, patients on ECMO need to be anticoagulated to avoid any thromboembolic complications and to guarantee proper pump function. One limitation of ECMO is the oxygenator, which serves a syphon for blood elements, particularly platelets, resulting in a diminution in oxygenator life-span. However, some of these issues may have been overcome with newer oxygenator designs as suggested by Pokersnik et al.10 As with all other devices mentioned herein, the major limitations to ECMO reside in cannulation site bleeding and limb complications.11 Nonetheless, we believe that when compared with other temporary MCS, ECMO is potentially easier to use and much more versatile as it may be applied to a variety of cardiopulmonary pathologies.

Finally, in this era of cost containment, it is important to understand differences between devices because if the outcomes associated to either support system are similar, then a difference may be found in cost. What the practitioner needs to understand is that each device has a capital cost (the amount of money needed to purchase the technology) and the cost of the disposables (the amount of money needed to replace the components for device after each use). The capital cost to invest in device technology is considerable and ranges between $35,000 and $50,000 per device console. These amounts do not include the cost of the disposable components, which can range from $1,000 to $32,000. As it may be inferred, devices such as the Impella carry a higher disposable price tag as each disposable contains the pump, which is at the center of this device technology. The TandemHeart, on the other hand, requires the purchase of special cannulas and pump as part of the disposable package. Conversely, the disposables used for ECMO are the least expensive of the three, but even when comparing manufacturers (i.e., Thoratec vs. MAQUET), there is a considerable difference in price for these components.12

Back to Top | Article Outline


There are few limitations in this cohort analysis. One limitation is a lack of power. The difference in proportions for bridge to transplant or long-term support and limb complications between the ECMO and mp-VAD groups was not statistically significant. A larger cohort of patients would be necessary to overcome this limitation. In this retrospective study, possible factors that may have biased the results were included. This study refers to a single-center database, and it is possible that selection of patients for a specific type of short-term MCS may differ among heart failure units across the country. Finally, there is some heterogeneity in our population because two types of percutaneous assist devices (Impella 5 LP and TandemHeart) were included. Overcoming these limitations will likely require a randomized trial with more homogeneous patient groups.

Back to Top | Article Outline


Similar outcomes can be expected when comparing mp-VAD with ECMO in patients with cardiogenic shock. Their use may be tailored to patient-specific cardiac etiologies, practitioner preference, and an understanding of cost.

Back to Top | Article Outline


The authors wish to thank Kelli R. Trungale, MLS, ELS, for editorial assistance.

Back to Top | Article Outline


1. Gibbon JH Jr. The development of the heart-lung apparatus. Am J Surg. 1978;135:608–619
2. Stoney WS. Evolution of cardiopulmonary bypass. Circulation. 2009;119:2844–2853
3. Cheng JM, den Uil CA, Hoeks SE, et al. Percutaneous left ventricular assist devices vs. intra-aortic balloon pump counterpulsation for treatment of cardiogenic shock: a meta-analysis of controlled trials. Eur Heart J. 2009;30:2102–2108
4. Thiele H, Sick P, Boudriot E, et al. Randomized comparison of intra-aortic balloon support with a percutaneous left ventricular assist device in patients with revascularized acute myocardial infarction complicated by cardiogenic shock. Eur Heart J. 2005;26:1276–1283
5. Lamarche Y, Cheung A, Ignaszewski A, et al. Comparative outcomes in cardiogenic shock patients managed with Impella microaxial pump or extracorporeal life support. J Thorac Cardiovasc Surg. 2011;142:60–65
6. Thiele H, Lauer B, Hambrecht R, Boudriot E, Cohen HA, Schuler G. Reversal of cardiogenic shock by percutaneous left atrial-to-femoral arterial bypass assistance. Circulation. 2001;104:2917–2922
7. Rajagopal V, Steahr G, Wilmer CI, Raval NY. A novel percutaneous mechanical biventricular bridge to recovery in severe cardiac allograft rejection. J Heart Lung Transplant. 2010;29:93–95
8. Kawashima D, Gojo S, Nishimura T, et al. Left ventricular mechanical support with Impella provides more ventricular unloading in heart failure than extracorporeal membrane oxygenation. ASAIO J. 2011;57:169–176
9. Meyns B, Dens J, Sergeant P, Herijgers P, Daenen W, Flameng W. Initial experiences with the Impella device in patients with cardiogenic shock - Impella support for cardiogenic shock. Thorac Cardiovasc Surg. 2003;51:312–317
10. Pokersnik JA, Buda T, Bashour CA, Gonzalez-Stawinski GV. Have changes in ECMO technology impacted outcomes in adult patients developing postcardiotomy cardiogenic shock? J Card Surg. 2012;27:246–252
11. Smedira NG, Moazami N, Golding CM, et al. Clinical experience with 202 adults receiving extracorporeal membrane oxygenation for cardiac failure: survival at five years. J Thorac Cardiovasc Surg. 2001;122:92–102
12. Yulong G, Xiaowei S, McCoach R, Kunselman A, El-Banayosy A, Undar A. Mechanical performance comparison between RotaFlow and CentriMag centrifugal blood pumps in an adult ECLS model. Perfusion. 2010;25:71–76

ECMO; Impella; mp-VAD; TandemHeart

Copyright © 2013 by the American Society for Artificial Internal Organs