(C-118) Flow fields in connector based vascular anastomosis

Introduction::

Vascular and reconstructive surgery often require vascular anastomosis, in which two severed blood vessels are reconnected to restore blood flow. The standard of care for anastomosis requires that the vessels be sutured around their periphery in such a way as to ensure continuous intimal contact. These procedures require extensive expertise and are highly time consuming even for expert surgeons. A possible approach that has been proposed for making these procedures more efficient is to insert a connector on the inner lumen of the two severed vessels, and lock this into place with a connector that passes around the outer diameter of the vessel (Figure 1). While substantially more efficient, the connector can cause perturbations to blood flow that lead to occlusion of the vessel connection by promoting thrombosis. The goals of this study were therefore to identify the features of flow that lead to the development of thrombosis, and to identify ways to possibly alleviate this complication.

Materials and Methods::

Numerical simulations were performed to estimate the vessel distension associated with insertion of a connector between the free ends of two severed blood vessels, and to assess how the connector and distension affected flow fields and the likelihood of thrombosis. Vessels were modeled as cylindrical, incompressible, isotropic, linear elastic tubes, and connectors as rigid and rotationally symmetric. The outer diameter of the connector was greater than the inner diameter of the vessel, causing distension to a shape that was estimated using axisymmetric finite element analysis. This shape served as input for axisymmetric computational fluid dynamics simulation of blood flow through the vessel and connector. Blood rheology was approximated using the Bird-Carreau model, which accounts for shear-thinning and viscoelasticity. The size and shape of the connector were studied parametrically. The inlet condition was defined by a centerline velocity. The outlet boundary condition was prescribed by the two element Windkessel model. The likelihood of thrombus formation was assessed using previously published models that account for the time and spatial extent of exposure of blood to strain rates that are pathologically high (strain rate > 1000 /s) and pathologically low (strain rate < 50 /s). Equations were solved in the COMSOL Multiphysics environment (version 6.1, COMSOL, Burlington, MA) using scripts in Matlab and Simulink (version 2023a, MathWorks, Natick, MA), and standard convergence studies were performed.

Results, Conclusions, and Discussions::

After analyzing blood flow patterns, pressure distributions, and wall stresses, regions of abnormal flow and potential sites of thrombus formation were identified. The presence of a connector between ends of the severed vessel led to elevations of shear rates at the leading corner of the connector, and to reduced shear rates near the wall at that edge (Figure 2).  These pathological flow fields could be attenuated by distending the vasculature sufficiently far.  Effects were independent of the length of the connector, and with the thickness of the vessel over the physiologically relevant range. Rounding the edges of the connector reduced peak shear strain rates. The outlet of the connector was associated with flow stagnation and low shear rates, but not with pathologically high shear rates. In all cases, thickening of the connector worsened flow fields, while distension of the vessel improved flow patterns up to a limit beyond which the inner surface of the connector lay outside of the original inner diameter of the vessel. These results provide guidance for the selection and design of vasculature connectors. 

Results demonstrate the significance of the choice of vascular connector on hemodynamics in vascular networks. Simulations revealed a strong negative effect of connector thickness, which can be attenuated by distension of the blood vessel.  Results also show a secondary effect of the sharpness of the leading edge of the connector, and no effect of the length of the connector.  Taken together, these results provide support for the feasibility of performing connector-based anastomosis and provide key guidance for the design of anastomosis connectors.

Acknowledgements (Optional): : The Center for Engineering Mechanobiology (CEMB), Dr. Guy Genin, Washington University in St. Louis, and the National Science Foundation (NSF)

References (Optional): :