Selected Talk: Cardiac Energetics and Systemic Perfusion with Veno-arterial Extracorporeal Membrane Oxygenation Versus ECMELLA for Cardiogenic Shock in a Large Animal Model

Permissions× For commercial reprint enquiries please contact Springer Healthcare:

For permissions and non-commercial reprint enquiries, please visit to start a request.

For author reprints, please email
Information image
Average (ratings)
No ratings
Your rating

Published online:

Support:The development of this supplement was funded by Abiomed.

Open Access:

This work is open access under the CC-BY-NC 4.0 License which allows users to copy, redistribute and make derivative works for non-commercial purposes, provided the original work is cited correctly.

Dr Banke’s research is based on patients with severe left ventricular (LV) dysfunction and cardiogenic shock (CS) after acute MI where mechanical circulatory support to provide end organ perfusion is necessary.

Pharmacological support or minor invasive support devices alone are often insufficient to provide end organ perfusion for these patients, and more robust circulatory support, such as veno-arterial extracorporeal membrane oxygenation (VA-ECMO), may be necessary. Nonetheless, VA-ECMO comes with a number of factors to consider. One such factor is the watershed, a phenomenon in which the blood coming from the ECMO flows in the opposite direction to blood coming from the LV. This has the potential of compromising the oxygen supply to the coronary arteries, and potentially the central nervous system and upper body. Another issue to consider is inadequate emptying of the LV during VA-ECMO support.1

Dr Banke explained that the current study aimed to address how LV unloading can be achieved in VA-ECMO-supported patients. Dr Banke’s study hypothesis was that the combination of VA-ECMO and Impella CP (ECMELLA) improves cardiac energetics without compromising end organ perfusion compared with VA-ECMO alone in a porcine model of CS. The model used was the Danish landrace pig, weighing approximately 70 kg and administered amiodarone prior to instrumentation to avoid ventricular arrhythmias. A Doppler flow probe was used around the carotid artery to collect flow data throughout the study period. VA-ECMO was accessed via the femoral artery and there was renal vein access and arterial access for coronary angiogram and placement of the Impella CP. A midline incision was made to allow for echocardiographic imaging during the study. Data collected included the pressure–volume area, stroke work and LV end-diastolic volume obtained from the conductance catheter and continuous cardiac output and venous oxygen saturation (SvO2) from the pulmonary artery catheter; in addition, arterial, renal vein and pulmonary artery blood was sampled for arterial blood gas analysis.

Following instrumentation, step-wise controlled CS was induced by embolisation in the left main coronary artery and was guided by SvO2.2 The definition of shock was at least a 50% reduction in cardiac output and/or at least a 50% reduction in SvO2 or an absolute SvO2 of <30%. Once CS had been induced, the animals were supported on VA-ECMO and underwent further embolisation until there was a decoupling of LV pressure and aortic pressure. Five animals entered the VA-ECMO arm, whereas seven animals entered the ECMELLA arm. Animals in both groups were monitored continuously for a further 4 hours on mechanical support.

Dr Banke presented the study results. The mean (±SD) arterial pressure in the VA-ECMO and ECMELLA groups after CS was 51 ± 5 mmHg and 51 ± 4 mmHg, respectively. After 4 hours, mean arterial pressure was lower in the VA-ECMO than ECMELLA group (64 ± 8 mmHg versus 69 ± 21 mmHg, respectively; p=0.02). There was no difference in noradrenaline infusion between the two groups. The VA-ECMO flow was equal at the time of CS but decreased in the ECMELLA group during the support period to 3 l/min, compared with 3.7 l/min in the VA-ECMO group (p=0.001). Total cardiac work, calculated as pressure–volume area × heart rate, was significantly lower in the ECMELLA group, with the greatest reduction in the first hour of mechanical support, after which it stabilised (p=0.003). End organ perfusion, measured as arterial lactate concentration, was equally elevated in both groups. There was no statistically significant difference in mixed SvO2 between the two groups, but SvO2 increased during the mechanical support period. Carotid blood flow increased during the mechanical support period by an equal amount of 100 ml in both groups. There was no difference in renal SvO2 between the two groups, but SvO2 increased during the first hour of support before stabilising. Urine output was numerically higher in the ECMELLA group, but the difference was not statistically significant (p=0.8).

Dr Banke concluded her presentation by summarising the key findings, which showed total cardiac work was lower in the ECMELLA compared with VA-ECMO group, indicating more appropriate cardiac energetics, while there was equal end organ perfusion in the two groups.


  1. Møller-Helgestad OK, Hyldebrandt JA, Banke A, et al. Impella CP or VA-ECMO in profound cardiogenic shock: left ventricular unloading and organ perfusion in a large animal model. EuroIntervention 2019;14:e1585–92. 
    Crossref | Pubmed
  2. Møller-Helgestad OK, Ravn HB, Møller JE. Large porcine model of profound acute ischemic cardiogenic shock. Methods Mol Biol 2018;1816:343–52. 
    Crossref | Pubmed