Right Heart Remodeling After Pulmonary Valve Replacement in Patients With Pulmonary Atresia or Critical Stenosis With Intact Ventricular Septum

Abstract

Background Patients with pulmonary atresia or critical pulmonary stenosis with intact ventricular septum (PA/IVS) and biventricular circulation may require pulmonary valve replacement (PVR). Right ventricular (RV) remodeling after PVR is well described in tetralogy of Fallot (TOF); we sought to investigate RV changes in PA/IVS using cardiac magnetic resonance imaging. Methods and Results A retrospective cohort of patients with PA/IVS who underwent PVR at Boston Children's Hospital from 1995 to 2021 with cardiac magnetic resonance imaging before and after PVR was matched 1:3 with patients with TOF by age at PVR. Median regression modeling was performed with post-PVR indexed RV end-diastolic volume as the primary outcome. A total of 20 patients with PA/IVS (cases) were matched with 60 patients with TOF (controls), with median age at PVR of 14 years. Pre-PVR indexed RV end-diastolic volume was similar between groups; cases had higher RV ejection fraction (51.4% versus 48.6%; P=0.03). Pre-PVR RV free wall and left ventricular (LV) longitudinal strain were similar, although LV midcavity circumferential strain was decreased in cases (-15.6 versus -17.1; P=0.001). At a median of 2 years after PVR, indexed RV end-diastolic volume was similarly reduced; cases continued to have higher RV ejection fraction (52.3% versus 46.9%; P=0.007) with less reduction in RV mass (Δ4.5 versus 9.6 g/m²; P=0.004). Post-PVR, RV and LV longitudinal strain remained unchanged, and LV circumferential strain was similar, although lower in cases. Conclusions Compared with patients with TOF, patients with PA/IVS demonstrate similar RV remodeling after PVR, with lower reduction in RV mass and comparatively higher RV ejection fraction. Although no differences were detected in peak systolic RV or LV strain values, further investigation of diastolic parameters is needed.

Keywords: congenital heart disease; pulmonary atresia; strain imaging; valve replacement.

Figure 1. Cardiac magnetic imaging strain measurement technique.

Cardiac magnetic resonance images in 2‐chamber short‐axis (A) and 4‐chamber long‐axis (B and C) views depict the border detection technique used in deriving LV and right ventricular strain. Longitudinal strain is derived from 4‐chamber views, and LV circumferential strain is derived from short‐axis views. Right‐hand panels depict the average strain (negative deflection) of the summated segments over 1 cardiac cycle. LV indicates left ventricular.

Figure 2. Volumetric and ventricular strain comparisons between PA/IVS and TOF by cardiac magnetic resonance.

Box and whiskers plots depict the RVEDVi (y axis) (A), RV EF (y axis) (B), RV long. strain (C), and LV long. strain (D) before and following pulmonary valve replacement for patients with PA/IVS (purple) and TOF (red). Boxes represent medians with whiskers for interquartile range. EF indicates ejection fraction; long, longitudinal; LV, left ventricular; PA/IVS, pulmonary atresia with intact ventricular septum; RV, right ventricular; RVEDVi, RV end‐diastolic volume indexed for body surface area; and TOF, tetralogy of Fallot.