International Journal of Cardiology Congenital Heart Disease (Aug 2021)
3D modeling and printing in large native right ventricle outflow tract to plan complex percutaneous pulmonary valve implantation
Abstract
Objective: Investigating accuracy of cardiac tomography (CT) derived post-processing3D reconstruction (CT-PPR) and 3D printing to predict percutaneous pulmonary valve implantation (PPVI) feasibility. Background: PPVI feasibility remains challenging in large native regurgitant right ventricle outflow tract (RVOT). Methods: Fifteen patients with large native RVOT were investigated. CT-PPR consisted in RVOT long-axis curvilinear reconstruction (LACR) to measure the landing zone (LZ), and 3D volume rendering for morphological evaluation. A STL was generated to create 3D printed model (flexible resin). Balloon sizing was subsequently performed to measure LZ diameter (3D-MBD), compared to invasive balloon diameter (IBD) during catheterization, considered as the Gold Standard. Two operators predicted the feasibility of PPVI using CT-PPR and 3D printed models independently and blinded to outcome. Results: On 3D printed models, RVOT shape was tubular in 5 patients, divergent in 5 patients, concave in 4 patients and convergent in one. Agreement with CT-PPR RVOT shape morphology assessment was observed in 93% of cases (Kappa coefficient 0.91, p < 0.0001). Minimal IBD was 26.0 [24.4–27.9] mm. Minimal LZ LACR diameters were well correlated to IBD (Spearman rho = 0.67, p = 0.007; r2 = 0.55, p = 0.002) with a mean underestimation bias of 2.8 mm. Minimal 3D-MBD was correlated to IBD (Spearman rho = 0.55, p = 0.04, r2 = 0.50, p = 0.003) with mean underestimation bias of 0.9 mm.PPVI was successful and uneventful in 11 patients, challenging in one and non-feasible in 3. Using CT-PPR and 3D printed models, interventionists predictions agreement with outcome was 93% and 87% (Kappa coefficient = 0.86, p = 0.0001, and Kappa coefficient = 0.65, p = 0.0007 respectively). Conclusion: RVOT 3D CT-PPR and printing allow reliable assessment of RVOT shape and LZ diameter prior to PPVI. Prediction of PPVI feasibility in challenging cases is facilitated.
