Percutaneous pulmonary valve implantation (PPVI) is a challenging intervention, especially in Tetralogy of Fallot (TOF) patients with large native outflow tract. We assess the ability of 3D printing to predict PPVI success.

We retrospectively printed 15cardiac models with flexible photopolymer resine from tomodensitometry. Patients (Median age 40 yo [26–56]) had TOF with large regurgitant native outflow tract. Especially patients with complex PPVI procedures were selected. Printing method was adjusted to mimic pulmonary artery elasticity. Balloon sizing was performed under fluoroscopy on the models and in patients. Models analysis were performed blinded to outcome.

Median minimal landing zone diameter was 25mm [22–27]. Two patients were referred for elective surgical valve replacement (outflow tract measured at 32mm after sizing in one and at 34mm after prestenting in the second). Sapien 3 valves were implanted in 13 patients after prestenting. Shape of the outflow tract was described as tubular (n=8), sandglass (n=1), chicane (n=2), funnel-shaped (n=2) and ballooning (n=1) on models. Correlation between invasive and models balloon calibration was not significant. Using Bland-Altman analysis, mean difference between the 2 techniques was 0.4±3.9mm (Fig. 1). Two patients were considered unsuitable for PPVI based on models. These 2 patients matched with the 2 referred for surgery. Assessing size and shape of the models, 2 interventionalists rated the expected complexity of PPVI from straightforward (n=11), to moderate (n=2) and complex (n=2). Analysis of outcome confirmed challenging procedures in 2 and uneventful in 12. Kappa agreement coefficient was 0.44, P=0.007.

3D printed cardiac models are useful to predict challenging PPVI but lacks accuracy to size the outflow tract. Prospective studies are needed to corroborate these preliminary results and better define the interest of this new tool before PPVI.