Objective Comparative ventricular arrhythmia (VA) outcomes following transcatheter (TC-PVR) or surgical pulmonary valve replacement (S-PVR) have not been evaluated. We sought to compare differences in VAs among patients with congenital heart disease (CHD) following TC-PVR or S-PVR.

Methods Patients with repaired CHD who underwent TC-PVR or S-PVR at the UCLA Medical Center from 2010 to 2016 were analysed retrospectively. Patients who underwent hybrid TC-PVR or had a diagnosis of congenitally corrected transposition of the great arteries were excluded. Patients were screened for a composite of non-intraoperative VA (the primary outcome variable), defined as symptomatic/recurrent non-sustained ventricular tachycardia (VT) requiring therapy, sustained VT or ventricular fibrillation. VA epochs were classified as 0–1 month (short-term), 1–12 months (mid-term) and ≥1 year (late-term).

Results Three hundred and two patients (TC-PVR, n=172 and S-PVR, n=130) were included. TC-PVR relative to S-PVR was associated with fewer clinically significant VAs in the first 30 days after valve implant (adjusted HR 0.20, p=0.002), but similar mid-term and late-term risks (adjusted HR 0.72, p=0.62 and adjusted HR 0.47, p=0.26, respectively). In propensity-adjusted models, S-PVR, patient age at PVR and native right ventricular outflow tract (RVOT) (vs bioprosthetic/conduit outflow tract) were independent predictors of early VA after pulmonary valve implantation (p<0.05 for all).

Conclusion Compared with S-PVR, TC-PVR was associated with reduced short-term but comparable mid-term and late-term VA burdens. Risk factors for VA after PVR included a surgical approach, valve implantation into a native RVOT and older age at PVR.