Minimized atrio-atrial extracorporeal membrane oxygenation: Feasibility study before planed mid-term lung assistance combined with right ventricular support

Objectives: Currently, the status of acute respiratory failure (ARF) with concomitant right ventricular failure (RVF) after cardiac surgical procedures shows a poor prognosis. We combined right ventricular unloading with extracorporeal oxygenation in an atrio-atrial extracorporeal membrane oxygenation (ECMO). The present experiment in an ovine model should prove the principal feasibility of our new approach and allow the characterization of the system during different circulatory conditions. Special focus was to analyze the influence of different cannula sizes and pump flows that allow providing an adequate oxygenation and decarboxilation.

Methods: Twelve sheep were cannulated by differently sized inflow cannulas (28Fr/20Fr) to the right atrium and outflow cannulas (25Fr/20Fr) through the lateral left atrial wall. Both were connected by a serial combination of a microaxial pump (Impella elect) and a membrane oxygenator (NovaLung). In the present substudy (n = 6), we focused on the effect of ECMO cannula size and flow rates on gas exchange during normal circulation (A: baseline), after apnoeic hypoxia (B: treated ARF) and increased right atrial afterload by pulmonary banding combined with apnoea (C: treated ARF+RHV). Hemodynamic monitoring and serial blood gas analysis was performed.

Results: On assist, no relevant hemodynamic changes during normal circulation (A). Respiratory parameters (PO2/PCO2), mean arterial pressure and right ventricular pressure normalized during phase B and C of experiment. Various cannular sizes showed no significant difference regarding oxygenation and decarboxylation. Higher flow rates of microaxial pump led to a significantly better oxygenation. (Exemplary: Treated ARF after apnoea (phase B):

20Fr/20Fr; 3.5 l/min flow: PO2 308 ± 36 mmHg – POC2 30 ± 2 mmHg

20Fr/20Fr; 2.0 l/min flow: PO2 143 ± 14 mmHg – POC2 39 ± 5 mmHg).

Conclusions: Our study demonstrates the feasibility of right heart support combined low resistance membrane oxygenator for additional effective lung support. The system – even with small cannular sizes – provides an effective oxygenation and decarboxylation, cause to right ventricular unloading and reduces the RV-afterload and treats RVF efficiently. Pump flow should be over 3.5 l/min. This could lead to the implementation of newly designed two stage cannulas in a trans-septal interventional approach. Pathologies (e.g. ischemic RHF and acute lung injury) will have to be evaluated in long time experiments.