Fontan hemodynamics from 100 patient-specific cardiac magnetic resonance studies: a computational fluid dynamics analysis

Abstract

Objectives: This study sought to quantify average hemodynamic metrics of the Fontan connection as reference for future investigations, compare connection types (intra-atrial vs extracardiac), and identify functional correlates using computational fluid dynamics in a large patient-specific cohort. Fontan hemodynamics, particularly power losses, are hypothesized to vary considerably among patients with a single ventricle and adversely affect systemic hemodynamics and ventricular function if suboptimal.

Methods: Fontan connection models were created from cardiac magnetic resonance scans for 100 patients. Phase velocity cardiac magnetic resonance in the aorta, vena cavae, and pulmonary arteries was used to prescribe patient-specific time-averaged flow boundary conditions for computational fluid dynamics with a customized, validated solver. Comparison with 4-dimensional cardiac magnetic resonance velocity data from selected patients was used to provide additional verification of simulations. Indexed Fontan power loss, connection resistance, and hepatic flow distribution were quantified and correlated with systemic patient characteristics.

Results: Indexed power loss varied by 2 orders of magnitude, whereas, on average, Fontan resistance was 15% to 20% of published values of pulmonary vascular resistance in single ventricles. A significant inverse relationship was observed between indexed power loss and both systemic venous flow and cardiac index. Comparison by connection type showed no differences between intra-atrial and extracardiac connections. Instead, the least efficient connections revealed adverse consequences from localized Fontan pathway stenosis.

Conclusions: Fontan power loss varies from patient to patient, and elevated levels are correlated with lower systemic flow and cardiac index. Fontan connection type does not influence hemodynamic efficiency, but an undersized or stenosed Fontan pathway or pulmonary arteries can be highly dissipative.

Figure 1.

3D velocity fields from time-averaged computational simulations (left) and representative phases from the CMR acquisitions (3 right panels) for 3 patients with an extracardiac connection. Comparison of hepatic distribution results (as %LPA) shown for each case. Arrow for Patient C highlights a conserved region of flow stagnation at the caval flow collision site.

Figure 2.

3D velocity fields between time-averaged computational simulations (left) and representative phases from the CMR acquisitions (3 right panels) for 3 patients with an intra-atrial connection. Comparison of hepatic distribution results (as %LPA) shown for each case.

Figure 3.

Bland-Altman mean vs. difference comparison of power losses derived from simulations using pulsatile and time-averaged boundary conditions. It is noted that the 95% confidence intervals are small compared to average power losses.

Figure 4.

Statistically significant correlations observed between A) HFD and total pulmonary flow distribution. B) iPL and BSA. C) iPL and Q_S. D) iPL and the Nakata Index.

Figure 5.

3D Velocity streamlines color coded by local velocity magnitude and vessel of origin (inset images: blue- IVC; red- SVC) for the 5 highest iPL connections.

Figure 6.

3D Velocity streamlines color coded by local velocity magnitude and vessel of origin (inset images: blue- IVC; red- SVC) for the 5 lowest iPL connections.