Estimation of pre-Fontan pulmonary vascular resistance in children with single ventricle heart disease at the Glenn stage: a multicenter study

The Fontan procedure, employed in the management of children with single ventricle congenital heart disease, continues to present long-term complications. Notably, certain complications associated with this procedure are linked to imbalances in the distribution of hepatic blood flow. One promising strategy to address this challenge involves employing a digital twin to simulate diverse Fontan configurations. The objective is to identify an optimal design that ensures balanced hepatic blood flow and minimizes power losses. However, successful implementation depends on accurate, patient-specific estimates of pulmonary vascular resistance (PVR) for each lung at the pre-Fontan (Glenn) stage. In clinical practice, only the total PVR is typically measured, via catheterization using the Fick principle, but individual lung resistances can be derived by combining pressure data from catheterization (Cath) with flow data from cardiac magnetic resonance imaging (CMR). Still, notable discrepancies exist: Fick-based total PVR often differs significantly from Cath-CMR-based PVR due to differences in flow quantification, and neither method can distinguish between proximal and distal resistances within the Glenn pathway. An alternative method for estimating PVR was previously developed using a computational fluid dynamics (CFD) optimization framework. This method demonstrated a favorable correlation with PVR estimates derived from Cath-CMR, although it was not directly compared to clinical PVR values derived using the Fick principle. In this study, we compare three methods for calculating PVR, namely Fick-based, Cath-CMR-based, and CFD-based, using patient data from three independent institutions. Our results show that Fick-based PVR values are, on average, significantly lower than those obtained via the Cath-CMR and CFD methods. The CFD-based total PVR estimates show good agreement with the total Cath-CMR-based PVR. However, the elevated left proximal resistance present in the CFD method leads to a significant underestimation of the left lung resistance by the Cath-CMR method. This underscores the significance of incorporating proximal resistance in PVR estimation and supports the potential utility of the CFD-based method for preoperative planning in single ventricle patients.