LAUSR

A 60-year old man developed cardiogenic shock after a complicated percutaneous coronary intervention. He initially required 4.0L/min peripheral veno-arterial extracorporeal membrane oxygenation (VA-ECMO) as mechanical circulatory support. An Impella CP was inserted at the femoral artery in a retrograde manner to reach the left ventricle by crossing the aortic valve under fluoroscopic guidance. It served to vent the left ventricle by drawing blood from it foward to the descending aorta. It was set at at slow revolution of P2 to achieve 1L/min forward flow, which was enough for left ventricular venting. His heart function gradually improved and we planned to step up the Impella support and wean off VA-ECMO. However, the patient developed recurrent ventricular tachycardia (VT) when the Impella was stepped up to P8 with 3 L/min flow while maintaining the 4 L/min VA-ECMO flow. Common differential diagnoses included new myocaradial infarction, stent thrombosis and other mechanical complications including ventricular septal rupture, free wall rupture or acute valvular dysfunction. Echocardiography showed none of these findings but that the left ventricle was collapsed, embracing the Impella (video 1 in ESM), and this left ventricular irritation by the Impella resulted in VT. The VA-ECMO flow was immediately decreased to 2L/min and the left ventricle refilled in few seconds (Fig. 1; video 2 in ESM), and the VT stopped. With various sources of venous return, complete drainage of left ventricle by Impella alone is not possible. In the combination of VA-ECMO, however, the right ventricle is also decompressed. Near complete left ventricular decompression is made possible. In our case, upon stepping up the Impella flow, the aortic diastolic pressure increased, closing the aortic valve throughout the cardiac cycle. The retrograde VA-ECMO flow could not fill the left ventricle. The left ventricle was decompressed by the Impella and the venous return was drained by the VA-ECMO. The Impella was working in a collapsed left ventricle, and this could result in left ventricular irritation, arrhythmia, malfunction of the Impella and even ventricular perforation. Upon stepping down the VA-ECMO flow, venous return increased and the left ventricle refilled, and the VT stopped. The interaction between VA-ECMO and Impella is complex. Upon adjustment of either support, we would recommend instantaneous echocardiography assessment on the Impella position and the changes in hemodynamics.