Effect of RVAD Cannulation Length on Right Ventricular Thrombosis Risk: An In Silico Investigation.

Left ventricular assist devices (LVADs) have been used off-label as long-term support of the right heart due to the lack of a clinically approved durable right VAD (RVAD). Whilst various techniques to reduce RVAD inflow cannula protrusion have been described, the implication of the protrusion length on right heart blood flow and subsequent risk of thrombosis remains poorly understood. This study investigates the influence of RVAD diaphragmatic cannulation length on right ventricular thrombosis risk using a patient-specific right ventricle in silico model validated with particle image velocimetry. Four cannulation lengths (5, 10, 15 and 25 mm) were evaluated in a one-way fluid-structure interaction simulation with boundary conditions generated from a lumped parameter model, simulating a biventricular supported condition. Simulation results demonstrated that the 25-mm cannulation length exhibited a lower thrombosis risk compared to 5-, 10- and 15-mm cannulation lengths due to improved flow energy distribution (25.2%, 24.4% and 17.8% increased), reduced stagnation volume (72%, 68% and 49% reduction), better washout rate (13.0%, 11.6% and 9.1% faster) and lower blood residence time (6% reduction). In the simulated scenario, our findings suggest that a longer RVAD diaphragmatic cannulation length may be beneficial in lowering thrombosis risk; however, further clinical studies are warranted.

In silico Prediction of Thrombosis Risk in a Ventricular Assist Device Supported Right Heart: The Impact of Cannulation Site.

Right ventricular assist device (RVAD) associated thrombosis is a serious complication that may arise due to unfavorable blood flow dynamics (blood stasis) caused by RVAD cannula protrusion within the chambers. This study aims to investigate the thrombosis risk of cannulation via the right atrium (RA) and right ventricle (RV) (diaphragmatic) under full RVAD support using computational fluid dynamics. A HeartWare HVAD inflow cannula was virtually implanted in either the RA or RV of a rigid-walled right heart geometry (including RA, RV, superior, and inferior vena cava) extracted from computed tomography data of a biventricular support patient. Transient simulations, validated with particle image velocimetry, were performed with constant inflow. Thrombosis risk was predicted by analyzing the time-averaged blood velocity, blood stagnation volume, washout rate, and blood residence time (BRT). Results showed that RA cannulation disturbed the physiological swirling flow structure which can be found in an uncannulated RA. This led to a large low-velocity recirculation flow in the RV, increasing the thrombosis risk. Contrarily, RV diaphragmatic cannulation showed better preservation of swirling flow in the RA and flow ejection into the RV. Consequently, RV diaphragmatic cannulation exhibited a better washout rate (99% vs. 57% of old blood was replaced in 12 s), lower blood stagnation volume (0.13 ml vs. 32.85 ml), and BRT (4.2 s vs. 7.1 s) than the RA cannulation in this simulated non-pulsatile case. Our findings suggest that RV diaphragmatic cannulation had a lower thrombosis risk and might be more favorable in a full RVAD-supported setting.