The study was prematurely stopped since enrollment was much slower than initially expected.

Veno-arterial extra-corporeal membrane oxygenation (VA-ECMO) is used as a rescue strategy for patients in acute hemodynamic deterioration such as cardiogenic shock and cardiopulmonary arrest with severe pulmonary congestion. VA ECMO is the fastest way to stabilize a patient with cardiogenic shock and improve end-organ perfusion. However, one of the major disadvantages of peripheral VA-ECMO is that it provides no left ventricular unloading and increases left ventricular (LV) afterload secondary to the retrograde blood flow. Therefore, LV wall tension and myocardial oxygen demand may actually increase in the setting of VA ECMO.

The Impella® device is a miniature rotary blood pump which can be inserted retrograde across the aortic valve. In this configuration, it withdraws blood from the LV and ejects it into the ascending aorta. It unloads the left ventricle, reducing LV wall tension and myocardial oxygen demand and increasing myocardial blood flow. The Impella® 5.0 is an FDA approved pump designed for intermediate support in patients with severe, cardiogenic shock. The axillary positioning allows for early extubation and ambulation and is more stable than groin placement.

In present practice, the decision to place an Impella® pump in VA-ECMO patients is based on the perceived need for direct LV unloading or when a bridge device is required to transition off ECMO support. Patients with peripheral VA ECMO are managed with inotropic agents at the beginning and once patients develop pulmonary edema mechanical LV unloading is considered electively. The advantage of LV unloading with Impella® has been demonstrated in recent studies. We also reported that concomitant implantation of Impella® with VA ECMO for LV unloading resulted in improved survival and recovery of ventricular performance in patients with cardiogenic shock. Compared to delayed elective LV unloading, early LV unloading could lead to decreased pulmonary edema, improved oxygenation delivery to the myocardium, increased chance of LV recovery and improved survival.

The objective of this prospective study is to assess whether the early direct ventricular unloading using axillary Impella® leads to higher rates of cardiac recovery, defined as survival free from mechanical circulatory support, heart transplantation or inotropic support at thirty days, compared with the conventional, elective placement of Impella® after developing significant pulmonary congestion.

This prospective, single-arm trial will include all consecutive patients undergoing cannulation of peripheral VA ECMO at the Massachusetts General Hospital (MGH) from April 2019 to March 2020. All enrolled patients will undergo surgical placement of Impella® via axillary artery within 48 hours after peripheral VA ECMO initiation and be managed with early extubation and ambulation strategy.

After Impella® placement, there will be no research procedures performed. Subjects will receive standard of care and the subject's progress will be documented throughout the 30 days study follow-up. The data obtained from the electronic medical record in Epic health record system at MGH will be reviewed throughout the study.

The primary outcome of this study will be survival at 30 days. Prespecified secondary end points will include the rate of death from cardiovascular causes, New York Heart Association (NYHA) functional class, LV function (assessed by echocardiography), and the rate of stroke, neurological functional status, acute kidney injury, vascular complications, and bleeding.

The enrolled subjects (early Impella®) would be compared with patients who underwent current elective placement of Impella® after cannulation of VA ECMO (elective Impella®) in the past two years. Student's t-test (for continuous variables) or Fisher exact test (for categorical variables) will be used for between-group comparisons. The primary endpoint in the data analysis is binary: 30-day survival. The null hypothesis of no difference between early Impella® and elective Impella® will be tested using logistic regression. The secondary endpoint, the rate of death from cardiovascular causes, NYHA functional class, LV function, and the rate of stroke, neurological functional status, acute kidney injury, vascular complications, and bleeding, will be addressed using competing risks hazard regression models. This is a pilot study for the future multicenter study. If this study demonstrates no inferiority or any significant superiority, we will proceed to do multicenter prospective study. Power analysis will be assessed based on the result of this preliminary study.

Inclusion Criteria:

• Age ≥18
• Impaired LV systolic function with ≤35% of left ventricular ejection fraction (LVEF)
• Enlarged LV with ≥50mm of left ventricular diastolic diameter (LVEDD) on echocardiogram

Exclusion Criteria:

• Non-cardiac etiology
• Surgically correctable cardiac abnormality
• Recent significant pulmonary embolism
• Severe pulmonary hypertension
• Acute aortic dissection
• Presence of mechanical aortic valve prosthesis
• Presence of left ventricle thrombus
• Pre-existing Impella® 5.0
• Critical aortic stenosis
• Uncorrectable system malperfusion under ECMO support
• Significant cerebrovascular accident

Neither identifiable nor non-identifiable data will be sent outside Partners or to Partners researchers not listed on the protocol.

Adverse event data were collected up to 30-days post VA-ECMO cannulation/Impella Insertion, whichever came first.

Each subject will be followed through 30 days post VA-ECMO cannulation/Impella Insertion. Adverse events data will be collected throughout this time period. Adverse events will be determined and assessed only through medical chart review.