Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 3.
Published in final edited form as: Circ Cardiovasc Imaging. 2024 Apr 3;17(4):e016104. doi: 10.1161/CIRCIMAGING.123.016104

Impact of Age-related change in Caval Flow Ratio on Hepatic Flow Distribution in the Fontan Circulation

Vijay Govindarajan 1,4,5, Lauren Marshall 1, Akshita Sahni 1, Maria Cetatoiu 1, Emily Eickhoff 1, Jocelyn Davee 1, Nicole St Clair 1, Noah Schulz 1, David M Hoganson 1,4, Peter E Hammer 1,4, Sunil Ghelani 1,2, Ashwin Prakash 1,2, Pedro J del Nido 1,2, Rahul H Rathod 2,3
PMCID: PMC11073583  NIHMSID: NIHMS1977892  PMID: 38567518

Abstract

Background:

The Fontan operation is a palliative technique for patients born with single ventricle heart disease. The superior vena cava (SVC), inferior vena cava (IVC), and hepatic veins are connected to the pulmonary arteries in a total cavopulmonary connection by an extracardiac (EC) conduit or a lateral tunnel (LT) connection. A balanced hepatic flow distribution (HFD) to both lungs is essential to prevent pulmonary arteriovenous malformations and cyanosis. HFD is highly dependent on the local hemodynamics. The effect of age-related changes in caval inflows on HFD was evaluated using cardiac magnetic resonance (CMR) data and patient-specific computational fluid dynamics (CFD) modeling.

Methods:

SVC and IVC flow from 414 Fontan patients were collected to establish a relationship between SVC:IVC flow ratio and age. CFD modeling was performed in 60 (30 EC and 30 LT) patient models to quantify the HFD that corresponded to patient ages of 3, 8, and 15 years, respectively.

Results:

SVC:IVC flow ratio inverted at ~8 years of age, indicating a clear shift to lower body flow predominance. Our data showed that variation of HFD in response to age-related changes in caval inflows (SVC:IVC = 2, 1, and 0.5 corresponded to ages 3, 8, and 15+ respectively) was not significant for EC but statistically significant for LT cohorts. For all three caval inflow ratios, a positive correlation existed between the IVC flow distribution to both the lungs and the HFD. However, as the SVC:IVC ratio changed from 2→0.5 (age 3→15+), the correlation’s strength decreased from 0.87→0.64, due to potential flow perturbation as IVC flow momentum increased.

Conclusion:

Our analysis provided quantitative insights into the impact of the changing caval inflows on Fontan’s long-term HFD, highlighting the importance of SVC:IVC variations over time on Fontan’s long-term hemodynamics. These findings broaden our understanding of Fontan hemodynamics and patient outcomes.

Subject Terms: Hemodynamics, Mechanisms, Congenital Heart Disease

Keywords: Fontan, hemodynamics, cardiac MRI, computational fluid dynamics, hepatic flow distribution

Graphical Abstract

graphic file with name nihms-1977892-f0001.jpg

Introduction

A successful Fontan procedure offers complete redirection of systemic venous return, improved arterial saturations, and decreased chronic ventricular volume overload, thereby improving longevity for many patients with single ventricle heart disease (SVHD)1. The Fontan circulation is the result of a total cavopulmonary connection between the superior vena cava (SVC), inferior vena cava (IVC), and the hepatic veins to the pulmonary arteries (PA) typically achieved either through an extracardiac (EC) conduit or a lateral tunnel (LT) connection. Although the Fontan procedure is lifesaving for patients with SVHD, this “unnatural” physiology can lead to serious early and late complications in many patients. These comorbidities include obligatory central venous hypertension, pulmonary arteriovenous malformations (PAVM) leading to cyanosis, ventricular dysfunction, arrhythmia, thrombus formation, stroke, protein losing enteropathy, exercise intolerance, pleural effusions, and hepatic and/or renal dysfunction 1–3.

PAVMs are an important cause of progressive cyanosis in patients with a Fontan circulation. PAVMs are abnormal communications that develop between the pulmonary artery and pulmonary veins, bypassing the normal capillary bed impacting gas exchange4 leading to progressive cyanosis and exercise intolerance4,5. Complete resolution of PAVMs has been achieved post-liver transplantation in the setting of severe liver disease, post-Fontan completion, and after redirection of hepatic venous blood to the affected lung6–9. These studies stress the importance of symmetric hepatic blood flow to both lungs for the prevention of PAVMs. Symmetric delivery of hepatic venous blood to each lung is dependent on local blood flow dynamics, Fontan geometries, and caval blood flow interactions10–13.

Cardiac magnetic resonance (CMR) imaging allows for the assessment of vascular anatomy and flow and, when combined with computational fluid dynamics (CFD), can offer patient-specific biomechanical insights improving our quantitative understanding of the complex hemodynamics in Fontan. Numerous CMR-CFD-based studies have been conducted to gain quantitative insights to evaluate Fontan blood flow dynamics14–20. These studies range from estimating the energy loss across the Fontan pathway15,18,21, determining optimal hemodynamics20, characterizing hepatic venous flow12, and predicting the non-Newtonian effects of blood in Fontan flow22. CFD-based modeling frameworks are now being extended to perform virtual surgeries to personalize and improve surgical predictability10,23–25. While these studies have offered important insights into Fontan hemodynamics, shortcomings included small sample sizes, exclusion of hepatic inflow resulting in a limited understanding of local flow dynamics in hepatic-IVC-Fontan confluence, and simulation of only single flow states.

The goal of this study is to perform a comprehensive analysis of hepatic flow distribution (HFD) in a large cohort of LT and EC Fontan patients and quantitatively evaluate the impact of growth-related changes in caval inflow on hemodynamics with particular emphasis on HFD. To this end, we aim to 1) describe the quantitative relationship between the change in SVC-to-IVC inflow (interchangeably referred to as caval inflow) ratio over age, 2) perform CMR-based patient CFD modeling at different SVC-to-IVC flow ratios to quantitatively evaluate the variability of HFD over time, and 3) evaluate the correlations between IVC flow distribution to branch PAs and HFD supported by momentum analysis to delineate the impact of increasing IVC flow on HFD.

Material and Methods

In addition to the data provided in the supplementary materials, any other data that support the findings of this study are available from the corresponding author upon reasonable request.

Patient selection – Inclusion and exclusion criteria

Patients after the Fontan operation who had a CMR at Boston Children’s Hospital from 2003 to 2023 were retrospectively reviewed. Inclusion criteria included patients who had an EC or LT Fontan procedure as their primary surgery with non-Fenestrated Fontan or status-post closure. Patients were excluded if they had an interrupted IVC, bilateral bidirectional Glenn procedure, prior history of PAVM, history of thrombosis, or aortopulmonary collateral flow > 40%. This retrospective analysis utilized anonymous and deidentified data from CMR imaging that was acquired as a part of routine clinical care and was approved by the institutional review board (IRB title: Utilizing Patient specific Computational Fluid Dynamic Modeling to Optimize Fontan procedure) at Boston Children’s Hospital with a waiver of informed consent.

A random subset of 30 patients with an EC Fontan and 30 patients with LT Fontan were chosen for computational flow modeling. To be included for CFD modeling, patients had to have minimal imaging artifacts, 3D isotropic steady-state free-precession imaging, and 2D flow assessment in the SVC, IVC, and branch PAs. Patient specific demographic information and CMR derived flow data are provided in Supplementary tables S1 and S2. Moreover, a subgroup comprising 26 patient models (13 EC and 13 LT) from the total 60 CFD models was selected to investigate the correlation between IVC and hepatic flow distributions to the lungs.

Patient-specific anatomy reconstruction from CMR images

Figure 1 illustrates the overall patient-specific workflow for evaluating Fontan flow dynamics for this study. The SVC, IVC, branch PAs, Fontan pathway, and major hepatic vessels were segmented using a threshold appropriate for the blood pool contrast from the 3D isotropic imaging sequences. Segmentation was performed using Mimics 25.0 (Materialise NV, Leuven, Belgium). The geometries were smoothed, and the mesh was constructed using 3-Matic Medical (Materialise NV, Leuven, Belgium) (Figure 1). The stereolithography (STL) mesh file was then imported into Fusion 360 (Autodesk, San Rafael, CA), converted to step format, and all vessels were trimmed.

Figure 1.

Figure 1

Open in a new tab

Workflow diagram for Computational fluid Dynamics (CFD)-based patient-specific flow analysis. (A) MRI-acquired patient-specific images are segmented and reconstructed to build an anatomically accurate 3-D model of the patient’s Fontan pathway; (B) An illustration of extracardiac and lateral tunnel Fontan model utilized in our flow modeling analysis; (C) Patient-specific CFD simulations are conducted with individually tuned lumped parameters to ensure that the model-predicted total flow distribution to the PAs is consistent with CMR-measured flow distribution (refer to Tables S1 and S2). For each patient model in our cohort, flow predictions were made for three caval inflow ratios: 2, 1, and 0.5; (D) The study analyzed predictions of HFD to the RPA and LPA, velocity, and momentum. SVC: Superior Vena Cava; IVC: Inferior Vena Cava; LPA: Left Pulmonary Artery; RPA: Right Pulmonary Artery; LA: Left Atrium, RL: Right Lungs; LL: Left Lungs; HFD: Hepatic Flow Distribution.

Computational flow modeling

All patient-specific CFD simulations were carried out using the commercial software ANSYS FLUENT V19.0 and 20.0 (ANSYS, Canonsburg, PA)26 with geometries meshed using ANSYS ICEM26. CFD model details such as mesh density (Figure S1) for adequate flow resolution, solver methodology that ensured solution convergence for each of these models are provided in the Supplementary Material.

The computational flow domain consisted of the patient’s Fontan pathway, SVC (and the bridging vein if present), left and right PAs, IVC, and the main hepatic veins that drained blood into the Fontan-IVC confluence. The hepatic vessels were included in the analysis to capture the realistic local flow dynamic changes at the hepatic-IVC-Fontan anastomosis as hepatic blood interacted with that from IVC at varying flow rates. The inlets (SVC, IVC, hepatic veins) and the outlets (PAs) were extended to ensure numerical stability during flow computation.

Velocities that corresponded to CMR-measured flow rates were directly applied at the SVC inlets. CMR-measured flow rates at the Fontan baffle were split between IVC and hepatic inflow with 25% of total flow from the hepatic vessels based on previous studies27. The PA outlets were defined as pressure boundary conditions with a lumped parameter model coupled at the outlets (Figure 1) that compute outlet pressure based on pulmonary vascular resistance (PVR). The pulmonary vascular resistance (PVR) values were individually tuned for each patient CFD model so that the predicted total flow distribution to the LPA and RPA were consistent with CMR measurements within 5% (Supplementary Tables S1 and S2). This ensured consistency between clinically measured total outflow and the model thus allowing for realistic capture of the inflow, outflow and the local flow dynamics within the Fontan pathway to the lungs. Quasi-steady simulations were performed to solve the governing equations (incompressible Navier-Stokes equation) with a Quemada non-Newtonian viscosity model28 of blood to capture the effect of hematocrit (45%) and the local shear rate (see Supplementary Materials). HFD and IVC flow distribution, to the lungs were quantified using massless point particles that were uniformly seeded at the hepatic vein and IVC inlets which were passively carried along with the flow to the PA outlets. These particles were tracked, and their flux quantified as they exited the PA outlets 13,23 (see Supplementary Material). For each patient model, CFD simulations based on CMR flow data were performed followed by two additional simulations to represent the age-based changes in SVC and IVC inflows.

Statistical Analysis

The Shapiro-Wilk test was performed to assess the normal distribution of data (See Supplementary data for details). We utilized the non-parametric Friedman’s chi square test for statistical significance among groups because the predicted data on hepatic flow distribution from our CFD analysis were not normally distributed and involved three time points (or caval inflow ratios) (See Supplementary Figure S2 for Q-Q plots). Friedman’s test is considered more powerful because it makes fewer assumptions than the parametric ANOVA for skewed distributions29. A P-value of less than 0.05 indicated a statistically significant difference in hepatic flow distribution between the three age groups. Finally, the Pearson correlation coefficient (R) was employed to establish the correlation between predicted IVC flow distribution to lungs and HFD for each caval inflow ratio with an assumption that an R > 0.5 represents a strong positive correlation (Detailed description of our statistical methodology is provided in the supplementary material).

RESULTS

Impact of age on caval inflows

There were 414 Fontan patients who met inclusion/exclusion criteria and comprised the main retrospective study group. The median age at CMR was 16 [IQR 12, 21] with a range from 2 to 40 years. Supplementary Table S1 details the demographic information of the entire study group. Figure 2A shows the SVC:IVC inflow ratios versus age. SVC:IVC inflow ratios of 2, 1, and 0.5 corresponded to approximate ages 3, 8, and 15 years old, respectively. There is a non-linear, exponential decline in SVC:IVC inflow ratio from ages 2 to 8, with equal SVC:IVC inflow at around age 8. After age 8, IVC flow predominates but plateaus around age 15 with an SVC:IVC ratio of around 0.5. Accordingly, three simulations were carried out per patient model to predict HFD variation at SVC:IVC ratios of 2, 1, and 0.5, also adjusting for an estimated age-based corresponding change in total flow30–32. Given that most Fontan procedures occur between ages 3–5 years, this change in SVC:IVC inflow ratio may explain some of the changes in HFD over time (see below). Figure 2B describes indexed total caval inflow (SVC + IVC) indexed to body surface area (BSA) versus age which tends to have less changes as patients get older.

Figure 2.

Figure 2.

Open in a new tab

SVC-IVC Analysis based on CMR data from 414 Fontan patients. (A) Shift in caval inflow ratio with age. Red dots indicate the flow ratios in which patient-specific CFD simulations were performed. Arrow indicates the typical time of Fontan completion in single ventricle patients; (B) Total caval inflow indexed to BSA vs. age. SVC: Superior Vena Cava; IVC: Inferior Vena Cava; BSA: Body Surface Area.

Impact of change in caval inflow on hepatic flow distribution

CFD predictions showed that variations of HFD in response to changes in caval inflow were different in both EC and LT cohorts. In patients with an EC Fontan (n=30), the median CFD-predicted differential pulmonary artery blood flow percentage to the right/left lungs was 54/46%, 53/47%, and 53/47% for caval SVC:IVC ratios of 2, 1, and 0.5 respectively (Figure 3A). Median differential HFD percentage to right/left lungs were 70/30%, 66/34%, and 65/35% for caval ratios of 2, 1, and 0.5, respectively (Figure 3B). There was no significant change in pulmonary artery blood flow distribution or HFD (P=0.46), with changes in age or SVC:IVC flow in patients with an EC Fontan conduit.

Figure 3.

Figure 3

Open in a new tab

Variation in total flow distribution and hepatic flow distribution in our Fontan cohort (A & B) Extracardiac (n=30) and (C & D) Lateral tunnel (n=30), as caval flow ratio shifts with age. Note that figure shows HFD to RPA (LPA = 1-RPA).* indicates statistically significant variation of HFD with age related changes in caval flow ratios for the LT Fontan cohort with p-value = 0.04 (Friedman’s chi square test). EC: Extracardiac Fontan type; LT; Lateral Tunnel Fontan Type; HFD: Hepatic Flow Distribution; RPA: Right Pulmonary Artery; SVC: Superior Vena Cava; IVC: Inferior Vena Cava.

For patients with a LT Fontan (n=30), CFD-predicted median differential pulmonary artery blood flow percentage to the right/left lungs was 58/42%, 58/42%, and 57/43% for caval ratios of 2, 1, and 0.5 respectively (Figure 3C). Median differential HFD percentage to right/ left lungs were 71/29%, 63/27%, and 61/39% for SVC:IVC caval ratios of 2, 1, and 0.5, respectively (Figure 3D). The change in HFD distribution across three caval inflow ratios was statistically significant (P=0.04). There were no other significant changes in pulmonary artery blood flow in patients with an LT Fontan conduit. These data suggest that the increased proportion of IVC (versus SVC) inflow are driving changes in HFD in patients with an LT Fontan, but not patients with an EC Fontan.

Strength of positive correlation between IVC flow distribution to PAs and HFD moderately decreases with age

Twenty-six patients (13 EC, 13 LT) randomly selected from our EC and LT cohorts were further analyzed to determine the correlation between IVC flow distribution to branch PAs and HFD to branch PAs. Strong positive correlations existed between HFD versus IVC flow distribution of r=0.86 (P=2.3e-08), r=0.73 (P=1.8e-05), and r=0.63 (P=5.5e-04) for SVC:IVC ratios of 2, 1, and 0.5, respectively (Figure 4). The strength of the correlation decreased slightly over increased IVC flow (or age). Specific patient examples of momentum analysis with graphical depictions are shown and described in Figures 5–8. The greater momentum of the SVC flow at age ~3 (left panels of Figures 5–8 A), seemed to preferentially divert the IVC flow that seemed to carry the hepatic blood flow (left panels of Figures 5–8 B and C). This flow phenomenon may be attributed to the strong positive correlation between HFD and IVC for SVC:IVC ratio of 2. As IVC flow increases with age (middle and right panels of Figures 5–8 A), flow disturbance induced by the increased momentum from IVC flow at the IVC-hepatic anastomosis and, eventually, competing with SVC at Fontan-PA-SVC anastomosis may have contributed to the slight decrease in correlation between HFD and IVC for SVC:IVC ratios of 2 and 0.5 (middle and right panels of Figures 5–8 B and C, Figure 4 B and C) .

Figure 4.

Figure 4

Open in a new tab

Correlation analysis between IVC flow distribution and HFD to RPA was performed using 26 patient CFD models (13 EC + 13 LT) to determine the Pearson correlation coefficient (ρ). (A) SVC:IVC flow ratio = 2; (B) SVC:IVC flow ratio = 1; (C) SVC:IVC flow ratio = 0.5. HFD: Hepatic Flow Distribution; RPA: Right Pulmonary Artery; SVC: Superior Vena Cava; IVC: Inferior Vena Cava.

Figure 5.

Figure 5

Open in a new tab

Momentum analysis on Patient 13 Extracardiac Fontan pathway. (A) Contour plots of normalized (to maximum) momentum; (B) % IVC flow distribution to lungs; and (C) % Hepatic flow distribution to lungs. SVC: Superior Vena Cava; IVC: Inferior Vena Cava.

Figure 8.

Figure 8

Open in a new tab

Momentum analysis on Patient 30 Lateral tunnel Fontan pathway (A): Contour plots of normalized (to maximum) momentum; (B) % IVC flow distribution to lungs; and (C) % Hepatic flow distribution to lungs. SVC: Superior Vena Cava; IVC: Inferior Vena Cava.

Discussion

Blood flow and its interaction with its lumen is a dynamic process that changes with any perturbation caused by either age, growth, disease, or intervention. These hemodynamic concepts are also relevant for the Fontan circulation. With the improved survival of Fontan patients leading to an increase in the worldwide population of surviving Fontan patients with various types of single ventricle anatomy5, a broader understanding of how Fontan hemodynamics alters with age can help with the management and prevention of future complications12,14–16,22,23. This study establishes the quantitative shift from SVC dominant to IVC dominant flow as Fontan patients age with an inversion point around age 8. In younger patients with a higher SVC:IVC ratio (2), both EC and LT Fontan patients had mild maldistribution of HFD (Figure 3A and C). Patients with a LT had a more favorable ‘correction’” with increased IVC flow (right HFD decreased from 71% → 61%) (Figure 3D). Patients with EC Fontan did not see this correction and remained steady across various flow conditions and with increasing age (Figure 3B). This study is the first to show that long-term variations in HFD are impacted by the type of Fontan procedure.

Results from this study show how SVC:IVC inflows into the Fontan pathway change with age and how these changes impact local hemodynamics and HFD. CMR data collected from a large (n=414) group of Fontan patients revealed a curvilinear change in SVC and IVC flow ratio with age (2/3rd SVC at age~3 transitioning to 1/3rd SVC at age 15+), indicating that the Fontan inflows can be dynamic at least until adulthood, upon which IVC flow dominates (Figure 2). These data are highly consistent with a previous Doppler study by Salim et al., that showed that SVC flow decreased (to 35%) as patients got older33. Our CFD model predictions showed that patients with a LT Fontan were more likely to have maldistribution of HFD when SVC inflow dominates (age ~3). This is clinically significant as imbalances to HFD can lead to PAVMs. This maldistribution seemed to improve with increased IVC flow. EC Fontan physiology was generally more favorable with more symmetric HFD early in life and more consistent even with changes in SVC:IVC flow ratios. Frieberg et al., have demonstrated a correlation between hepatic to pulmonary blood flow balance and oxygen saturation, even in the absence of PAVMs34. Such correlation helps establish the significant link between CFD predictions and clinical findings. In this context, our findings suggest that as hepatic to pulmonary blood flow balance changes over time, there may be a corresponding alteration in oxygen saturations as patients grow. However, a focused study is warranted to establish correlation that will be conducted in the future. Our data reinforce the importance of growth-related changes in physiology that should be incorporated when predicting any long-term Fontan hemodynamics.

While it is desirable to have a balanced HFD to prevent PAVMs in Fontan patients, a quantitatively established minimum level of hepatic factor required to prevent PAVMs is currently lacking. Several studies have demonstrated that a complete lack of hepatic blood effluence can lead to the formation of PAVMs35–37. At the same time, there is clinical evidence showing complete resolution of PAVM following redirection of hepatic venous blood to the affected lung6,7,11,38. In our patient cohort, while many patients had a balanced HFD (Figure 3), some patient model predictions showed highly skewed HFD especially when SVC inflow dominated the Fontan inflows (age ~3). Our model predictions suggested that HFD tends to achieve some balance with age for these patients consistent with other patient models. These data highlight patient-specificity, how SVC-IVC caval inflow changes with respect to age can impact HFD, and the need for patient-specific planning of Fontan pathways.

Long term outcomes of extracardiac versus lateral tunnel Fontan procedures has been a subject of continuous debate5,15, with most decision making highly influenced by individual surgeon choice and institutional preferences5. A comprehensive evaluation of power loss utilizing patient-specific CFD modeling by Haggerty et al15, revealed no considerable differences between the LT and EC connection types. The current study provides new quantitative insights with respect to HFD and introduces some concern for early maldistribution of HFD in both Fontan types, but improved balance in patients with a LT Fontan.

There is a growing interest in and utilization of patient-specific computational fluid dynamics to tailor the design and enhance the Fontan pathway, aiming to maximize the distribution of hepatic flow while minimizing energy loss 10,23–25,34. In these studies, neither hepatic vein geometry nor the hepatic flow is explicitly modeled; instead, HFD is quantified assuming that hepatic flow and IVC flow are thoroughly mixed. Results from our study where hepatic vein geometry was incorporated in all of our patient model to capture the realistic streaming of hepatic flow as they obliquely meet the IVC (Figure 1, 5–8) indicate that the correlation between IVC flow distribution and HFD to lungs tend to diminish with age (R = 0.86→0.63; Figure 4). These findings could carry significant clinical implications, emphasizing the importance of integrating hepatic vein geometries in CFD-based surgical planning, particularly for older patients undergoing Fontan revision or upsizing to allow realistic predictions of HFD to lungs. Finally, workflow introduced in this study incorporating the caval flow shift, can be added to an existing CFD-based Fontan surgical planning to predict the robustness of a planned Fontan pathway with respect to long-term delivery of HFD to lungs.

Dasi et al identified the caval-offset induced momentum barrier created by the SVC flow being the main factor for unbalanced HFD in extracardiac Fontan12. These findings are highly consistent with our results, especially for age ~3 when SVC flow dominated (Figures. 5–8). Also, the highest skewness in HFD occurred for an SVC-IVC flow ratio of 2 (Age ~3) (Figure. 3) for both EC and LT cohorts highlighting the impact of SVC. Furthermore, our momentum analysis revealed that the potential long-term impact of the SVC momentum barrier, while significant immediately after Fontan completion (Age ~3) diminishes as IVC flow dominated with patient growth (age 8 and 15+, Figures 5–8). These results demonstrate how competing momentum between SVC and IVC flow in the Fontan pathway impacts HFD during the course of time.

Our study has the following limitations. Variability in SVC:IVC flow measurements for younger patients in our study may be attributed to the relatively lesser density of patient data in that group and to the fluctuation in cerebral blood flow (CBF) induced by general anesthesia39, commonly used for image quality assurance in younger patients during CMR studies. For example, anesthetic agents like nitrous oxide and ketamine increase CBF39–42, while xenon and propofol decrease it39,43,44. The difference in sedation protocols between younger patients and older children/adults during CMR may have contributed to the observed variation in SVC:IVC flow for younger patients. All the simulations were performed using time-averaged flow conditions measured from the MRI with rigid wall assumptions. Rigid wall assumptions were made based on the conclusions from previous study that there was no significant change in HFD due to wall deformation45. The growth of PAs was not accounted for based on the study that reported nonsignificant growth of vessels commensurate with increasing flow21,46. Patient-specific simulations such as those presented in our studies that incorporate vessels of different sizes and tortuosity require superior image quality and reconstruction that are made possible by improvements in CMR protocols and segmentation workflows. Fontan flow is known to be transient (e.g.: impact of respiration) and hence, the time-average flow conditions may introduce small quantitative differences. The potential growth and geometry size changes of the LT baffle that can occur with time were also not able to be estimated in this study. Finally, a constant PVR was assumed for all the SVC-to-IVC caval flow ratios, while PVR can potentially change over time.

In conclusion, this study described the SVC:IVC flow ratios in a large cohort of patients ranging from 2 to 40 years old, demonstrating the transition from SVC dominant flow to IVC dominant flow at approximately 8 years of age. Using CMR-based patient-specific CFD modeling, both EC and LT Fontan patients had a tendency towards maldistribution of HFD early in line. With increasing IVC inflow with age, patients with a LT had improved HFD distribution, while EC Fontan patients had a relatively consistent HFD. Our analysis provided quantitative insights into the impact of changing caval inflows on long-term hepatic flow distribution in extracardiac and lateral tunnel Fontan types. These data also showed the importance of incorporating SVC:IVC changes over time in CFD modeling toward gaining insights into long-term hemodynamics of Fontan.

Supplementary Material

Supplemental Publication Material

Figure 6.

Figure 6

Open in a new tab

Momentum analysis on Patient 16 Extracardiac Fontan pathway. (A) Contour plots of normalized (to maximum) momentum; (B) % IVC flow distribution to lungs; and (C) % Hepatic flow distribution to lungs. SVC: Superior Vena Cava; IVC: Inferior Vena Cava.

Figure 7.

Figure 7

Open in a new tab

Momentum analysis on Patient 13 Lateral tunnel Fontan pathway. (A): Contour plots of normalized (to maximum) momentum; (B) % IVC flow distribution to lungs; and (C) % Hepatic flow distribution to lungs. SVC: Superior Vena Cava; IVC: Inferior Vena Cava.

Clinical Perspective.

With improvement in standard of care and management of single ventricle patients with Fontan physiology, the population of adults with Fontan circulation is increasing. Consequently, there is a clinical need to comprehend the impact of patient growth on Fontan hemodynamics. Using CMR data, we were able to quantify the relationship between changing caval inflows and somatic growth. We then used patient-specific computational flow modeling to quantify how this relationship affected the distribution of long-term hepatic flow in extracardiac and lateral tunnel Fontan types and quantified how IVC flow distribution correlated with hepatic flow distribution to lungs. Our findings demonstrated the significance of including SVC:IVC changes over time in CFD modeling to learn more about the long-term hemodynamics of Fontan. Fontan surgical approaches are increasingly planned and optimized using computational flow modeling. For a patient undergoing a Fontan procedure, the workflow presented in this study that takes into account the variations in Caval inflows over time can provide predictive estimates of the long-term hemodynamics in a planned Fontan pathway.

Acknowledgements:

Research reported in this publication was supported by the National Heart, Lung, And Blood Institute of the National Institutes of Health under Award Number R01HL161507. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Authors also acknowledge the computational resources provided by the Texas Advanced Computer Center for performing the flow simulations as a part of this study.

Funding:

Research reported in this publication was supported by the National Heart, Lung, And Blood Institute of the National Institutes of Health under Award Number R01HL161507 and in-kind research support from NSF (high-performance computing resources from Texas Advanced Computer Center).

Abbreviations

SVHD

Single Ventricle Heart Disease

IVC

Inferior Vena Cava

SVC

Superior Vena Cava

PA

Pulmonary artery

PAVM

Pulmonary Arteriovenous Malformations

TCPC

Total Cavopulmonary Connection

HFD

Hepatic Flow Distribution

EC

Extracardiac conduit

LT

Lateral Tunnel

PVR

Pulmonary Vascular Resistance

CMR

Cardiac Magnetic Resonance

CFD

Computational Fluid Dynamics

CBF

Cerebral Blood Flow

Footnotes

Supplementary Data

Supplementary Methods contains detailed description of methodology used in computational flow modeling and statistical analysis.

Disclosure

Funding Disclosure/Conflict of Interest Statement: VG reports research funding from the AHA (19TPA34860013), the NHLBI/NIH (R01HL161507) and in-kind research support from NSF (high-performance computing resources from TACC). AS, EE, DH, PEH, and RR report research funding from the NHLBI/NIH (R01HL161507). In addition, VG reports collaborative research funding from Polyvascular Corp, Houston, TX.

Supplemental Material:

Supplemental Methods, Figures S1–S2, Tables S1–S4, References 47–52

References

  • 1.Mazza GA, Gribaudo E, Agnoletti G. The pathophysiology and complications of Fontan circulation. Acta Bio Medica: Atenei Parmensis 2021;92: e2021260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Kavarana MN, Jones JA, Stroud RE, Bradley SM, Ikonomidis JS, Mukherjee R. Pulmonary arteriovenous malformations after the superior cavopulmonary shunt: mechanisms and clinical implications. Expert Rev Cardiovasc Ther 2014;12:703–713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Maxey TS, Herlong JR, Jansen LN, Kirshbom PM. Fontan Procedure. In: Critical Heart Disease in Infants and Children. Elsevier; 2019:747–757. e742. [Google Scholar]
  • 4.Cartin-Ceba R, Swanson KL, Krowka MJ. Pulmonary arteriovenous malformations. Chest. 2013;144:1033–1044. [DOI] [PubMed] [Google Scholar]
  • 5.Rychik J, Atz AM, Celermajer DS, Deal BJ, Gatzoulis MA, Gewillig MH, Hsia T-Y, Hsu DT, Kovacs AH, McCrindle BW. Evaluation and management of the child and adult with Fontan circulation: a scientific statement from the American Heart Association. Circulation. 2019;140:e234–e284. [DOI] [PubMed] [Google Scholar]
  • 6.Pike NA, Vricella LA, Feinstein JA, Black MD, Reitz BA. Regression of severe pulmonary arteriovenous malformations after Fontan revision and “hepatic factor” rerouting. Ann Thorac Surg 2004;78:697–699. [DOI] [PubMed] [Google Scholar]
  • 7.Wu I-H, Nguyen KH. Redirection of hepatic drainage for treatment of pulmonary arteriovenous malformations following the Fontan procedure. Pediatr Cardiol 2006;27:519–522. [DOI] [PubMed] [Google Scholar]
  • 8.Srivastava D, Preminger T, Lock JE, Mandell V, Keane JF, Mayer JE Jr, Kozakewich H, Spevak PJ. Hepatic venous blood and the development of pulmonary arteriovenous malformations in congenital heart disease. Circulation. 1995;92:1217–1222. [DOI] [PubMed] [Google Scholar]
  • 9.Spearman AD, Ginde S. Pulmonary Vascular Sequelae of Palliated Single Ventricle Circulation: Arteriovenous Malformations and Aortopulmonary Collaterals. J Cardiovasc Dev Dis 2022;9:309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Yang W, Vignon-Clementel IE, Troianowski G, Reddy VM, Feinstein JA, Marsden AL. Hepatic blood flow distribution and performance in conventional and novel Y-graft Fontan geometries: a case series computational fluid dynamics study. J Thorac Cardiovasc Surg 2012;143:1086–1097. [DOI] [PubMed] [Google Scholar]
  • 11.Sundareswaran KS, de Zélicourt D, Sharma S, Kanter KR, Spray TL, Rossignac J, Sotiropoulos F, Fogel MA, Yoganathan AP. Correction of pulmonary arteriovenous malformation using image-based surgical planning. JACC Cardiovasc Imaging. 2009;2:1024–1030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Dasi LP, Whitehead K, Pekkan K, de Zelicourt D, Sundareswaran K, Kanter K, Fogel MA, Yoganathan AP. Pulmonary hepatic flow distribution in total cavopulmonary connections: extracardiac versus intracardiac. J Thorac Cardiovasc Surg 2011;141:207–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.De Zélicourt DA, Haggerty CM, Sundareswaran KS, Whited BS, Rossignac JR, Kanter KR, Gaynor JW, Spray TL, Sotiropoulos F, Fogel MA. Individualized computer-based surgical planning to address pulmonary arteriovenous malformations in patients with a single ventricle with an interrupted inferior vena cava and azygous continuation. J Thorac Cardiovasc Surg 2011;141:1170–1177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Whitehead KK, Pekkan K, Kitajima HD, Paridon SM, Yoganathan AP, Fogel MA. Nonlinear power loss during exercise in single-ventricle patients after the Fontan: insights from computational fluid dynamics. Circulation. 2007;116:I-165-I-171. [DOI] [PubMed] [Google Scholar]
  • 15.Haggerty CM, Restrepo M, Tang E, de Zélicourt DA, Sundareswaran KS, Mirabella L, Bethel J, Whitehead KK, Fogel MA, Yoganathan AP. Fontan hemodynamics from 100 patient-specific cardiac magnetic resonance studies: a computational fluid dynamics analysis. J Thorac Cardiovasc Surg 2014;148:1481–1489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Khiabani RH, Whitehead KK, Han D, Restrepo M, Tang E, Bethel J, Paridon SM, Fogel MA, Yoganathan AP. Exercise capacity in single-ventricle patients after Fontan correlates with haemodynamic energy loss in TCPC. Heart. 2015;101:139–143. [DOI] [PubMed] [Google Scholar]
  • 17.Wang Q, Primiano C, McKay R, Kodali S, Sun W. CT image-based engineering analysis of transcatheter aortic valve replacement. JACC Cardiovasc Imaging. 2014;7:526–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Bossers SS, Cibis M, Gijsen FJ, Schokking M, Strengers JL, Verhaart RF, Moelker A, Wentzel JJ, Helbing WA. Computational fluid dynamics in Fontan patients to evaluate power loss during simulated exercise. Heart. 2014;100:696–701. [DOI] [PubMed] [Google Scholar]
  • 19.Cibis M, Jarvis K, Markl M, Rose M, Rigsby C, Barker AJ, Wentzel JJ. The effect of resolution on viscous dissipation measured with 4D flow MRI in patients with Fontan circulation: Evaluation using computational fluid dynamics. J Biomech 2015;48:2984–2989. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Bove E, De Leval M, Migliavacca F, Balossino R, Dubini G. Toward optimal hemodynamics: computer modeling of the Fontan circuit. Pediatr Cardiol 2007;28:477–481. [DOI] [PubMed] [Google Scholar]
  • 21.Restrepo M, Tang E, Haggerty CM, Khiabani RH, Mirabella L, Bethel J, Valente AM, Whitehead KK, McElhinney DB, Fogel MA. Energetic implications of vessel growth and flow changes over time in Fontan patients. Ann Thorac Surg 2015;99:163–170. [DOI] [PubMed] [Google Scholar]
  • 22.Wei Z, Singh-Gryzbon S, Trusty PM, Huddleston C, Zhang Y, Fogel MA, Veneziani A, Yoganathan AP. Non-Newtonian effects on patient-specific modeling of Fontan hemodynamics. Ann Biomed Eng 2020;48:2204–2217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Trusty PM, Wei ZA, Slesnick TC, Kanter KR, Spray TL, Fogel MA, Yoganathan AP. The first cohort of prospective Fontan surgical planning patients with follow-up data: How accurate is surgical planning? J Thorac Cardiovasc Surg 2019;157:1146–1155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.van Bakel TM, Lau KD, Hirsch-Romano J, Trimarchi S, Dorfman AL, Figueroa CA. Patient-specific modeling of hemodynamics: supporting surgical planning in a Fontan circulation correction. J Cardiovasc Transl Res 2018;11:145–155. [DOI] [PubMed] [Google Scholar]
  • 25.Frieberg P, Aristokleous N, Sjöberg P, Töger J, Liuba P, Carlsson M. Computational fluid dynamics support for fontan planning in minutes, not hours: the next step in clinical pre-interventional simulations. J Cardiovasc Transl Res 2022;15:708–720. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.ANSYS F ANSYS fluent theory guide 19.1. ANSYS, Canonsburg, PA 2019. [Google Scholar]
  • 27.Eipel C, Abshagen K, Vollmar B. Regulation of hepatic blood flow: the hepatic arterial buffer response revisited. World J Gastroenterol 2010;16:6046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Quemada D Rheology of concentrated disperse systems II. A model for non-newtonian shear viscosity in steady flows. Rheologica Acta 1978;17:632–642. [Google Scholar]
  • 29.Pereira DG, Afonso A, Medeiros FM. Overview of Friedman’s test and post-hoc analysis. Comm Statist Simulation Comput 2015;44:2636–2653. [Google Scholar]
  • 30.Cattermole GN, Leung PM, Mak PS, Chan SS, Graham CA, Rainer TH. The normal ranges of cardiovascular parameters in children measured using the Ultrasonic Cardiac Output Monitor. Crit Care Med 2010;38:1875–1881. [DOI] [PubMed] [Google Scholar]
  • 31.Yoganathan AP, Fogel M, Gamble S, Morton M, Schmidt P, Secunda J, Vidmar S, Del Nido P. A new paradigm for obtaining marketing approval for pediatric-sized prosthetic heart valves. J Thorac Cardiovasc Surg 2013;146:879–886. [DOI] [PubMed] [Google Scholar]
  • 32.Mannarino S, Bulzomì P, Codazzi AC, Rispoli GA, Tinelli C, De Silvestri A, Manzoni F, Chiapedi S. Inferior vena cava, abdominal aorta, and IVC-to-aorta ratio in healthy Caucasian children: Ultrasound Z-scores according to BSA and age. J Cardiol 2019;74:388–393. [DOI] [PubMed] [Google Scholar]
  • 33.Salim MA, DiSessa TG, Arheart KL, Alpert BS. Contribution of superior vena caval flow to total cardiac output in children: a Doppler echocardiographic study. Circulation. 1995;92:1860–1865. [DOI] [PubMed] [Google Scholar]
  • 34.Frieberg P, Sjöberg P, Hedström E, Carlsson M, Liuba P. In vivo hepatic flow distribution by computational fluid dynamics can predict pulmonary flow distribution in patients with Fontan circulation. Sci Rep 2023;13:18206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Shah MJ, Rychik J, Fogel MA, Murphy JD, Jacobs ML. Pulmonary AV malformations after superior cavopulmonary connection: resolution after inclusion of hepatic veins in the pulmonary circulation. Ann Thorac Surg 1997;63:960–963. [DOI] [PubMed] [Google Scholar]
  • 36.Ashrafian H, Swan L. The mechanism of formation of pulmonary arteriovenous malformations associated with the classic Glenn shunt (superior cavopulmonary anastomosis). Heart. 2002;88:639–639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Vettukattil J Pathogenesis of pulmonary arteriovenous malformations: role of hepatopulmonary interactions. Heart. 2002;88:561–563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.McElhinney DB, Marx GR, Marshall AC, Mayer JE, Del Nido PJ. Cavopulmonary pathway modification in patients with heterotaxy and newly diagnosed or persistent pulmonary arteriovenous malformations after a modified Fontan operation. J Thorac Cardiovasc Surg 2011;141:1362–1370. e1361. [DOI] [PubMed] [Google Scholar]
  • 39.Slupe AM, Kirsch JR. Effects of anesthesia on cerebral blood flow, metabolism, and neuroprotection. J Cereb Blood Flow Metab 2018;38:2192–2208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Wollman H, Alexander SC, Cohen PJ, Smith TC, Chase PE, van der Molen RA. Cerebral circulation during general anesthesia and hyperventilation in man. Paper/Poster presented at: ASA; 1965; [DOI] [PubMed] [Google Scholar]
  • 41.Leon JE, Bissonnette B. Transcranial Doppler sonography: nitrous oxide and cerebral blood flow velocity in children. Can J Anaesth 1991;38:974–979. [DOI] [PubMed] [Google Scholar]
  • 42.Field L, Dorrance D, Krzeminska E, Barsoum L. Effect of nitrous oxide on cerebral blood flow in normal humans. Br J Anaesth 1993;70:154–159. [DOI] [PubMed] [Google Scholar]
  • 43.Kaisti KK, Långsjö JW, Aalto S, Oikonen V, Sipilä H, Teräs M, Hinkka S, Metsähonkala L, Scheinin H. Effects of sevoflurane, propofol, and adjunct nitrous oxide on regional cerebral blood flow, oxygen consumption, and blood volume in humans. Anesthesiology. 2003;99:603–613. [DOI] [PubMed] [Google Scholar]
  • 44.Schlünzen L, Juul N, Hansen K, Cold G. Regional cerebral blood flow and glucose metabolism during propofol anaesthesia in healthy subjects studied with positron emission tomography. Acta Anaesthesiol Scand 2012;56:248–255. [DOI] [PubMed] [Google Scholar]
  • 45.Long C, Hsu MC, Bazilevs Y, Feinstein J, Marsden A. Fluid–structure interaction simulations of the Fontan procedure using variable wall properties. Int J Numer Method Biomed Eng 2012;28:513–527. [DOI] [PubMed] [Google Scholar]
  • 46.Restrepo M, Mirabella L, Tang E, Haggerty CM, Khiabani RH, Fynn-Thompson F, Valente AM, McElhinney DB, Fogel MA, Yoganathan AP. Fontan pathway growth: a quantitative evaluation of lateral tunnel and extracardiac cavopulmonary connections using serial cardiac magnetic resonance. Ann Thorac Surg 2014;97:916–922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Quemada D A rheological model for studying the hematocrit dependence of red cell-red cell and red cell-protein interactions in blood. Biorheology. 1981;18:501–516. [DOI] [PubMed] [Google Scholar]
  • 48.Abbasian M, Shams M, Valizadeh Z, Moshfegh A, Javadzadegan A and Cheng S. Effects of different non-Newtonian models on unsteady blood flow hemodynamics in patient-specific arterial models with in-vivo validation. Comput Methods Programs Biomed 2020;186:105185. [DOI] [PubMed] [Google Scholar]
  • 49.Issa RI, Gosman A and Watkins A. The computation of compressible and incompressible recirculating flows by a non-iterative implicit scheme. J Comput Phys 1986;62:66–82. [Google Scholar]
  • 50.Barrett R, Berry M, Chan TF, Demmel J, Donato J, Dongarra J, Eijkhout V, Pozo R, Romine C and Van der Vorst H. Templates for the solution of linear systems: building blocks for iterative methods: SIAM; 1994. [Google Scholar]
  • 51.Lawrence K ANSYS Tutorial Release 2020: SDC Publications; 2020 [Google Scholar]
  • 52.Spiegel M, Redel T, Zhang YJ, Struffert T, Hornegger J, Grossman RG, Doerfler A and Karmonik C. Tetrahedral vs. polyhedral mesh size evaluation on flow velocity and wall shear stress for cerebral hemodynamic simulation. Comput Methods Biomech Biomed Engin 2011;14:9–22. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Publication Material
NIHMS1977892-supplement-Supplemental_Publication_Material.docx (2.8MB, docx)

RESOURCES