By Van Selby, MD

Assistant Professor of Medicine, University of California, San Francisco Cardiology Division, Advanced Heart Failure Section

Dr. Selby reports he is a consultant for Alnylam Pharmaceuticals and Akcea Therapeutics.

SYNOPSIS: In a series of patients with cardiogenic shock, the use of venoarterial extracorporeal membrane oxygenation with a percutaneous left ventricular unloading device was associated with better-than-expected outcomes.

SOURCE: Schrage B, Burkhoff D, Rübsamen N, et al. Unloading of the left ventricle during venoarterial extracorporeal membrane oxygenation therapy in cardiogenic shock. JACC Heart Fail 2018;6:1035-1043.

For patients with severe cardiogenic shock (CS), venoarterial extracorporeal membrane oxygenation (VA-ECMO) is used to restore end-organ perfusion. However, one drawback of VA-ECMO is the associated increase in left ventricular (LV) afterload, which leads to multiple negative consequences in patients with severely impaired cardiac function. To minimize the effects of increased LV afterload, many practitioners implement tactics to unload the LV in patients supported with the VA-ECMO method. One approach is with the use of a percutaneous transaortic LV assist device such as the Impella. The results of smaller studies have suggested the addition of an Impella to VA-ECMO (an approach referred to as “ECMELLA”) may improve outcomes compared to VA-ECMO alone.

Schrage et al studied consecutive patients who were treated with the ECMELLA approach for cardiogenic shock. LV unloading was performed using an Impella (either 2.5 or CP) placed femorally. The primary outcome was 30-day all-cause mortality. Since there was no control group, the authors compared the observed mortality to what would be expected based on two previously validated risk models: the Survival After Venoarterial ECMO (SAVE) and Simplified Acute Physiology (SAPS-II) scores.

Between 2013 and January 2018, 106 patients were treated with ECMELLA. The most common indication was CS due to myocardial infarction. At 30 days of follow-up, all-cause survival was 35.8% compared to expected survival of 20% and 6.9% based on the median SAVE and SAPS-II scores, respectively. Successful weaning from VA-ECMO was achieved in 51.9% of patients, and 11.3% went on to receive a durable LV assist device.

Right heart catheterization data were available from the implant procedure in three patients. All three cases showed an acute rise in pulmonary capillary wedge pressure (PCWP) after initiation of VA-ECMO, followed by a rapid decline in PCWP once the Impella was placed. The authors concluded that implantation of a transaortic LV assist device in combination with VA-ECMO was associated with better-than-predicted outcomes in patients with cardiogenic shock.


Use of VA-ECMO is expanding rapidly, both for the treatment of severe cardiogenic shock and extracorporeal cardiopulmonary resuscitation (eCPR). However, despite significant progress in the technology and understanding behind VA-ECMO, survival remains poor. While there are many factors contributing to poor outcomes in this critically ill population, one potentially addressable issue is the significant increase in LV afterload that occurs because of retrograde perfusion of the aorta by the arterial ECMO cannula. In patients with severely compromised cardiac function, the LV is unable to overcome this afterload and eject enough blood into the aorta.

As venous return to the LV continues, there is a rise in LV filling pressures that leads to increased LV wall stress and myocardial oxygen demand, impairing myocardial recovery. Also, the rising LV filling pressure transmits to the pulmonary capillary bed, worsening pulmonary edema. In severe cases during which the aortic valve no longer opens, stasis of blood can lead to LV thrombus formation.

Based on these observations, the practice of LV venting, or tactics to unload the ventricle, have gained popularity. One key to improving outcomes for VA-ECMO patients is the generation of high-quality clinical research data to help refine both patient selection and management. Here, Schrage et al reported the largest series to date on ECMELLA. They demonstrated that ECMELLA is feasible in patients with severe CS and results in a clear, rapid improvement in left heart filling pressures. Furthermore, the authors suggested that ECMELLA is associated with substantial improvements in all-cause survival. However, because this was a nonrandomized, single-arm study, the only comparison was the expected survival based on two previously validated risk scores. This comparison is filled with caveats and should not be considered definitive for establishing the superiority of ECMELLA over VA-ECMO alone.

Considering the study design, there remain many unanswered questions. First, which patients require LV venting? Schrage et al used Impella in all VA-ECMO patients; therefore, the study design cannot answer this. The authors of many small studies have found that not all VA-ECMO patients require LV venting. Given the high cost and potential complications associated with placement of an Impella, it is important that we use it only when necessary.

On a similar note, this study also cannot tell us the best technique for venting the LV. While ECMELLA is one of the more popular options, there are many others, including intra-aortic balloon pumps, inotropic agents, vasodilators, and atrial septostomy. There are theoretical advantages to ECMELLA over many of these approaches, but the study design does not allow us to evaluate this important question.

Hopefully, randomized trials comparing various methods for LV venting will help clarify these important questions going forward. For now, studies such as this one are crucial as we seek to improve the care of patients treated with VA-ECMO.

While providers will continue to debate which patients to vent and how, Schrage et al have at least demonstrated that ECMELLA is feasible and successfully reduces left-sided pressures in patients with severe cardiogenic shock.