Structured Weaning From the Impella Left Ventricular Micro-Axial Pump in Acute Myocardial Infarction With Cardiogenic Shock and Protected Percutaneous Coronary Intervention: Experience From a Non-Cardiac Surgical Centre.

BACKGROUND: The Impella (Abiomed, Danvers, MA, USA) temporary percutaneous left ventricular assist device is increasingly used as mechanical circulatory support in patients with acute myocardial infarction-cardiogenic shock (AMICS) or those undergoing high-risk protected percutaneous coronary intervention (PCI). The optimal weaning regimen remains to be defined.

METHOD: We implemented a structured weaning protocol in a series of 10 consecutive patients receiving Impella support for protected PCI or AMICS treated with PCI in a high volume non-cardiac surgery centre. Weaning after revascularisation was titrated to native heart recovery using both haemodynamic and echocardiographic parameters.

RESULTS: Ten patients (eight male, two female; aged 43-70 years) received Impella support for AMICS (80%) or protected PCI (20%). Cardiogenic shock was of Society for Cardiac Angiography & Interventions grade C-E of severity in 80%, and median left ventricular end-diastolic pressure was 31 mmHg. Protocol implementation allowed successful weaning in eight of 10 patients with a median support time of 29 hours (range, 4-48 hours). Explantation was associated with an increase in heart rate (81 vs 88 bpm; p=0.005), but no significant change in Cardiac Index (2.9 vs 2.9 L/min/m2 ), mean arterial pressure (79 vs 82 mmHg), vasopressor requirement (10% vs 10%), or serum lactate (1.0 vs 1.0). Median durations of intensive care and hospital stay were 3 and 6 days, respectively. At 30 days, the mortality rate was 20%, with median left ventricular ejection fraction of 40%.

CONCLUSIONS: A structured and dynamic weaning protocol for patients with AMICS and protected PCI supported by the Impella device is feasible in a non-cardiac surgery centre. Larger studies are needed to assess generalisability of such a weaning protocol.