BIO16: Computation Fluid Dynamics Analysis of Dual Lumen VV ECMO Cannulas

Background: Venovenous extracorporeal membrane oxygenation (VV ECMO) treatment requires blood to be drained from the patient’s venous system, oxygenated in an artificial membrane, and returned. Traditionally, this has been achieved with two, single lumen cannulas (SLCs). More recently, dual lumen cannulas (DLCs) combining both drainage and return functions into a single device have become increasingly popular. Recirculation fraction, the proportion of oxygenated return blood which is directly drained out before passing through the tricuspid valve, should be low. Another key design objective for these cannulas is to minimize shear stress, as high levels increase the risk of coagulation activation (hemolysis). Whilst DLCs are thought to have superior recirculation performance compared to SLCs, this has not yet been demonstrated in a direct comparison. Furthermore, the shear stress produced by DLCs, their impact on native hemodynamics and response to changes in positioning have not been reported in adults. In this study, we aim to describe the recirculation and hemodynamic performance of DLCs. Methods: A patient-averaged model of the right atrium (RA) and the venae cavae was used to model venous flow. Two common DLCs, the Maquet Avalon Elite (Getinge) and MC3 Crescent (Medtronic) were converted into 3D models (Figure 1A.). Both were downscaled to 27Fr for our patient model. To delineate these from the off-the-shelf products we use the prefix “ds” for downscaled. Both DLCs were first positioned optimally and simulated at 2,4 and 6 L/min. Noting a high degree of similarity, insertion depth (±4cm) and rotation (±30° and ±60°) was only assessed for the dsAvalon.Figure 1. A. Downscaled (ds)Avalon and dsCrescent cannula features. B. Drainage fractions from the superior (SVC) and inferior vena cava (IVC) C. Time-averaged velocity streamlines of the high-velocity jet exiting the reinfusion port of both DLCs at 4L/min. Results: Pressure drop curves were nearly identical for the two devices. dsAvalon pressure drops agreed well with experimental data. Recirculation fraction was low for both DLCs compared to SLCs. SVC and IVC drainage adapted with ECMO flow rate to draw a greater proportion of flow from the IVC at high ECMO flow rates (Figure 1B). The reinfusion port design created a high-velocity jet of blood (Figure 1C). Consequently, time-averaged wall shear stress (TAWSS) was very high (>475 Pa) at the reinfusion port for both designs. Discussion: DLCs exhibit superior recirculation performance compared to SLCs under ideal conditions. This appears to be due to DLCs ability to drain flow from both the IVC and SVC, favoring greater IVC drainage at higher ECMO flow rates. Both designs produce a highly focused reinfusion jet which may increase risk of hemolysis and coagulation activation.