Computational Modeling In Design Of Endovascular Chemofilter Device For Removing Toxins From Blood

Summary form only given. Intra-arterial chemotherapy delivery (IAC) for treating cancer can lead to significant cardiac toxicity due to the drainage of excess drug to the systemic circulation. A catheter-based Chemofilter device temporarily inserted into the veins downstream of the tumor can remove chemotherapy drugs out of the blood stream right after these drugs have had their effect on the tumor. In this research, computational modeling is used to design the Chemofilter and optimize its hemodynamics performance. The Chemofilter's membrane is formed by a lattice of microcells loaded with immobilized DNA. The filter can sequester the chemo drugs, i.e. Doxorubicin (Dox) with intrinsic DNA binding characteristics, from flowing blood by absorbing Dox on its surface. A multiscale approach is used to analyze the local flow through the 100 micron cells of the membrane, as well as through the whole device deployed in the inferior vena cava. The Navier-Stokes and Darcy's flow equations are solved numerically, using computational fluid dynamics (CFD) solver ANSYS Fluent. The size of the microcells as well as the thickness and shape of the membrane are parameterized in the CFD simulations in order to reduce the pressure drop across the membrane while increasing the drug absorption. To calculate the filter's porous properties, the flow through a 2x2 matrix of microcells is modeled with periodic boundary conditions for different orientations relative to the flow and for various microcell sizes and number of layers. In the next step, the obtained permeability coefficients are used to model whole device as a thin porous membrane. Alternative Chemofilter geometries, including the paraboloid, cone, and umbrella shapes are simulated. The results show that the membrane can be modeled as an isotropic material. The pressure drop increases with increasing the number of microcell layers, while the porosity and permeability are increasing. For a membrane consisting of two microcell layers, increasing the cell size from 100 to 200μm resulted in reducing the pressure drop from 626 to 322 Pa. The results show that computational models can be invaluable for designing cardiovascular devices, optimizing drug absorption and reducing thrombosis, while minimizing animal testing.