Computational simulation-derived hemodynamic and biomechanical properties of the pulmonary arterial tree early in the course of ventricular septal defects

Abstract

Untreated ventricular septal defects (VSDs) can lead to pulmonary arterial hypertension (PAH) characterized by elevated pulmonary artery (PA) pressure and vascular remodeling, known as PAH associated with congenital heart disease (PAH-CHD). Though previous studies have investigated hemodynamic effects on vascular mechanobiology in late-stage PAH, hemodynamics leading to PAH-CHD initiation have not been fully quantified. We hypothesize that abnormal hemodynamics from left-to-right shunting in early stage VSDs affects PA biomechanical properties leading to PAH initiation. To model PA hemodynamics in healthy, small, moderate, and large VSD conditions prior to the onset of vascular remodeling, computational fluid dynamics simulations were performed using a 3D finite element model of a healthy 1-year-old's proximal PAs and a body-surface-area-scaled 0D distal PA tree. VSD conditions were modeled with increased pulmonary blood flow to represent degrees of left-to-right shunting. In the proximal PAs, pressure, flow, strain, and wall shear stress (WSS) increased with increasing VSD size; oscillatory shear index decreased with increasing VSD size in the larger PA vessels. WSS was higher in smaller diameter vessels and increased with VSD size, with the large VSD condition exhibiting WSS >100 dyn/cm[Formula: see text], well above values typically used to study dysfunctional mechanotransduction pathways in PAH. This study is the first to estimate hemodynamic and biomechanical metrics in the entire pediatric PA tree with VSD severity at the stage leading to PAH initiation and has implications for future studies assessing effects of abnormal mechanical stimuli on endothelial cells and vascular wall mechanics that occur during PAH-CHD initiation and progression.

Keywords: Computational cardiovascular biomechanics; Congenital heart defect; Pulmonary arterial hypertension; Pulmonary arterial tree; Ventricular septal defect.

Fig. 1

Simulation set up for the 3D PA model with a flow waveform prescribed at the MPA inlet and 3-element Windkessel models prescribed at every outlet to represent the downstream vascular dynamics. Fluid-structure interaction of PA wall deformability was modeled using the coupled momentum method, and external tissue support was prescribed on the walls via Robin boundary conditions representing springs and dashpots

Fig. 2

Optimization of distal PA morphometric tree incorporating a resistance of 3D PA model, b resistance calculation of distal PA tree with Poiseuille assumptions, c Nelder-Mead optimization of target distal PA tree resistances per order, and d optimization of the scaling constants for the diameter and lengths of the PA tree defining the geometry and informed by BSA scaling laws, 3D outlet diameters, and morphometric PA data

Fig. 3

Hemodynamics of the distal 0D PA tree were calculated from extrapolated flows from the 3D PA simulation and optimized geometry (illustrated) using Poiseuille assumptions

Fig. 4

The 3D hemodynamic simulations show a the pressure (blue) and flow (red) in the MPA increasing with increased VSD severity, b the spatially and temporally averaged pressure in all segments increasing with VSD severity, and c the average flow in all segments increased with a power law of diameter (represented by solid lines) for all VSD conditions

Fig. 5

The 3D proximal PA simulations show a increasing amplitudes in the time-dependent area waveform of the main PA (MPA) with VSD severity and b increases in the strain with VSD severity for each diameter-defined Strahler order identified in the 3D PA model. Order 12 corresponds to the most distal and smallest vessels in the PA model, and order 16 represents the MPA. (*) denotes significant differences with p-values < 0.05 between normal and all VSD conditions for orders 12-15

Fig. 6

Oscillatory shear index (OSI) in the 3D proximal PA simulations generally decrease with VSD severity in the larger vessel orders representing segmental, interlobar, and the main PAs. (*) denotes significance with p-value < 0.05, (**) denotes significance with p-value < 0.01

Fig. 7

Hemodynamic simulations of the 3D proximal PA model show that a the temporally and spatially averaged wall shear stress (WSS) in all 3D PA vessels increased with VSD severity with significant differences across all VSD conditions. b For the main PA (MPA), left PA (LPA), and right PA (RPA), the WSS increased with VSD severity and had a marked increase in value from MPA to the LPA and RPA lobar branches. c The time-averaged WSS magnitude in the 3D proximal PA models shows dynamic changes throughout the PA tree with higher WSS in the smaller vessels and in the larger VSD conditions

Fig. 8

Simulations in the distal PA morphometric tree using a 0D model and the proximal 3D PA model show increases in WSS magnitude with VSD severity grouped by vessel orders against vessel diameter. The WSS in the 0D distal PA model increased down the PA tree as vessel diameters decreased