Characterizing Aortic Arch Morphology in Pre- and Post-Fontan Hypoplastic Left Heart Syndrome Patients Using Statistical Shape Modeling (RF_FR_424)

Presenting Author(s)

• RO

  Ryan P. O'Hara, PhD

  Postdoc
  Children's National Hospital

Primary Author(s)

• RO

  Ryan P. O'Hara, PhD

  Postdoc
  Children's National Hospital

Co-Author(s)

• YS

  Yuv Singh, N/A

  Research Assistant
  Children's National Hospital
• JS

  John Sullivan, MD

  Fellow
  Children's National Hospital
• FC

  Francesco Capuano, PhD

  Assistant Professor
  Universitat Politecnica deCatalunya Barcelona Tech, Spain
• 

  Yue-Hin Loke, MD

  Associate Professor
  Children's National Medical Center

Background: Hypoplastic left heart syndrome (HLHS) is a congenital heart defect characterized by underdevelopment of the left heart structures. The aortic arch in patients with HLHS undergoes significant remodeling due to altered hemodynamics and surgical interventions. Aortic arch remodeling can indicate impaired ventricular function and lead to long-term complications throughout patient development. Statistical shape modeling (SSM) has proven advantageous in analyzing anatomical variations in other cardiovascular applications. However, previous studies in patients with HLHS have primarily focused on remodeling of the ventricle. The objective of this study is to perform SSM of the aortic arch in patients with HLHS to characterize morphological variations and their associations with clinical parameters and outcomes. 

Methods: In this retrospective single-center study, cardiovascular magnetic resonance (CMR) images were obtained in 41 patients with HLHS (27 Fontan, 14 pre-Fontan, Table 1). 3D models of the aorta were segmented and reconstructed with Slicer 3D. The ascending and descending aortic segments were clipped along the plane of the RVOT and apex of the heart, respectively. SSM was performed via ShapeWorks Studio (NIH/NIGMS CIBC) to derive the mean templates and quantified shape deviations (shape mode) of the Fontan and pre-Fontan aortic reconstructions. Principal component analysis was applied to extract the primary modes of shape variation and analyze their relationships to cardiac parameters and clinical variables.

Results: SSM captured over 85% of the total shape variance in the first 5 modes of the Fontan cohort and first 4 modes of the pre-Fontan cohort. The mean shape template in the Fontan group was represented by a widened ascending aorta and consistent descending aorta. The mean shape template in the pre-Fontan group was represented by narrowing of the aortic neck, bulging of the arch, and narrowing of the descending aorta (Figure 1). In the Fontan group, Mode 1 showed significant positive correlations with RV EDV_i and RV Mass_i, while Mode 4 correlated with LV EDV_i, shown in Table 2. In the pre-Fontan group, Mode 2 exhibited a strong correlation with GCS, and Mode 4 showed strong correlations with LV EDV_i and LV Mass_i, but without statistical significance. Fontan pressure had a moderate positive correlation with Mode 5. In the pre-Fontan group, Glenn pressure correlated significantly with Mode 3 (Table 2).

Conclusion: SSM enables comprehensive 3D quantification of aortic arch morphology in patients with HLHS. Strain measurements did not show significant correlations with shape modes in either cohort. The differences in correlations between Fontan and pre-Fontan groups indicate that aortic morphology and its relationship to cardiac hemodynamic function changes following the Fontan procedure. SSM of the aortic arch may provide patient-specific metrics to define HLHS pathology and predict adverse outcomes for HLHS patients.

Shape Modes of Variation in Fontan and Pre-Fontan Aortic Arch Reconstruction