Abstract Body (Do not enter title and authors here): Introduction
Right ventricular (RV) function is critical to the long-term health of patients with hypoplastic left heart syndrome (HLHS). However, conventional measurements do not consider the range of RV morphology that may impact RV function and lead to adverse outcomes. In this study, we determined associations between RV shape and clinical outcomes of HLHS using statistical shape modeling (SSM) and a large cardiac magnetic resonance (CMR) multicenter registry.

Methods
Three hundred and twenty-eight studies from unique HLHS patients in the Fontan Outcomes Registry using CMR Examinations (FORCE) were post-processed by a core lab at one institution, including volumetry and strain analysis. 3D end-diastolic models of RV were reconstructed using Mimics (Materialise). SSM was performed via Shapeworks Studio (NIH/NIGMS CIBC) to derive mean template and quantified shape variations (shape mode) representing deviation from the mean. RV shape relationship with mortality/heart transplant was assessed using bivariable/multivariable analyses with stepwise model selection.

Results
The mean template from all 328 patients resembled a circumferentially dilated RV with loss of concavity in septal wall (Fig A). Several generalized variants were identified (Fig B), particularly p1 – a sigmoidal / “apical bulge” variant where the RV apex dilates and "wraps” around the hypoplastic left ventricle (Fig C). Twenty-eight patients (8.5%) experienced mortality/heart transplant at 2.5±2.4 years after CMR. On bivariable analysis, p1 was associated with mortality/heart transplant, along with larger indexed RV end-diastolic volume/mass and reduced global circumferential strain. On multivariable analysis, p1 and indexed RV mass were associated with mortality/heart transplant (Table).

Conclusions
Statistical Shape Modeling can define the “usual” RV shape of HLHS. A sigmoidal shape variant is associated with adverse outcomes. SSM may provide patient-specific metrics to define RV dysfunction in HLHS.