Abstract 14364: Hemodynamic Evaluation of 30 Y-Graft Fontan Completions

Abstract

Background: Fontan completion, resulting in a total cavopulmonary connection (TCPC), is accomplished using either a traditional baffle or bifurcated Y-graft. The use of Y-grafts was originally hypothesized to provide more even hepatic blood flow distribution to the lungs, a factor related to pulmonary arteriovenous malformations. This study evaluates the hemodynamic performance of the largest Y-graft cohort to date, compares those results with a traditional baffle, and highlights Y-graft cases that exhibit both favorable and poor performance. Methods: Thirty Y-graft and 30 traditional baffle Fontan patients were analyzed. TCPC anatomies and vessel flow waveforms were reconstructed using cardiac magnetic resonance (CMR) images and phase-contrast CMR. Computational fluid dynamic simulations were performed to quantify TCPC power loss, resistance, and hepatic flow distribution (HFD). Comparisons between graft types and subanalysis of favorable versus poor performing Y-grafts were investigated. Results: The TCPC resistance distribution, HFD distribution, and the effect of LPA stenosis on power loss for Y-grafts and traditional baffles are shown in Fig. 1A-1C. TCPC resistance was significantly lower in traditional baffle patients. HFD was similar overall, with median and range of 40 (1-77) and 51 (1-99) for the traditional baffle and Y-graft, but show a discrepancy at extreme values with more unbalanced flow in the Y-graft cohort. Power loss was more sensitive to LPA stenosis in the Y-graft cohort. Prediction of Y-graft HFD is multi-factorial and six representative cases are shown in Fig. 1D. Conclusions: Y-grafts do not inherently provide more balanced HFD than traditional baffles. Traditional baffles are more energetically favorable and less sensitive to PA stenosis. Graft type should be considered on an individual basis as hemodynamic performance is based on a combination of factors, including global flow distribution, PA stenosis and SVC positioning.