Objective

To investigate the impact of veno-venous extracorporeal membrane oxygenation (VV-ECMO) cannula configuration and ECMO blood flow on intravascular hemodynamic environment, oxygen saturation, hemolysis, and thrombotic risk under varying perfusion conditions.

Methods

Four different VV-ECMO cannulation configurations (femoro-jugular double cannulation, femoro-femoral double cannulation, dual-lumen single bicaval cannulation, and dual-lumen single bicaval cannulation blocking the atrial connecting vessel) commonly used in clinical practice were modeled and analyzed under different ECMO flow support conditions using computational fluid dynamics method. Oxygen saturation and flow rate at the ECMO outlet were used to determine the recirculation fraction (RF). The red blood cell damage model was used to explore the hemolysis risk. A thrombosis model considering platelet activation and deposition was proposed to quantify platelet activation and thrombosis risk.

Results

Different cannula configurations and ECMO flowrates significantly affect the RF, proximal wall vascular injury, and hemolysis risk in VV-ECMO. At low ECMO flow rates (≤3 L/min), all cannula configurations exhibit reduced recirculation and blood injury, while oxygen saturation at the pulmonary artery outlet is lower. Higher ECMO flows increase recirculation and shear stress while enhancing oxygen delivery. The use of the dual-lumen cannula instead of two-cannula configuration can simultaneously minimize recirculation and blood vascular injury while achieving higher oxygen saturation in the pulmonary artery under the same ECMO flow rate condition. The effect of drainage cannula side-hole blockage on the surrounding hemodynamic environment differs significantly across cannulation configurations. Side-hole blockage of the drainage cannula had the greatest impact on femoro-jugular double cannulations, with less hemodynamic impact on other cannula configurations, especially in double-lumen cannulation. In femoro-jugular double cannulas, side-hole blockage leads to increased RF and decreased pulmonary oxygen saturation, resulting in reduced effective oxygen supply from ECMO, while it also reduces vessel wall shear stress and hemolysis risk.

Conclusion

For VV-ECMO, the dual-lumen cannula configuration maintains lower recirculation and blood damage while ensuring adequate oxygen supply, outperforming two-cannula configurations. Among the two-cannula configurations, femoro-femoral cannulation yields the lowest recirculation but the highest shear-related injury. Pulmonary artery oxygen content is a more reliable indicator of VV-ECMO oxygenation capacity than RF. Notably, there exists a high risk of thrombosis in proximity to the drainage cannula, and the blockage of the drainage cannula increases the ECMO RF and decreases the oxygenation efficiency. These findings provide guidance for selecting and designing VV-ECMO cannula configurations.