Multi-institutional assessment of practice patterns in the use of cardiovascular magnetic resonance and computed tomography for pulmonary valve replacement planning: A survey-based study

BACKGROUND: Consensus documents provide recommendations on using cardiovascular magnetic resonance (CMR) and cardiovascular computed tomography (CCT) in pulmonary valve replacement (PVR) planning in congenital heart disease (CHD). However, clinical practice patterns across institutions remain unclear, particularly with the increase in transcatheter PVR options and multi-modality imaging use. We aimed to characterize practice variation in cross-sectional imaging using CCT and CMR. METHODS: A REDCap survey was distributed to Society for Cardiovascular Magnetic Resonance and Society of Cardiovascular Computed Tomography members to assess imaging practices for PVR planning in CHD, including imaging volumes, acquisition protocols, right ventricular outflow tract (RVOT) measurement techniques, and reporting times, with one response per institution. Wilcoxon rank sum tests, Fisher's exact tests, and analysis of variance were used. RESULTS: Forty-three institutions (84% US and 16% international) responded. Most centers (29/43, 67%) reported performing CMR first, adding CCT only if the Harmony valve or Alterra pre-stent was under consideration. Only two centers (5%) used CCT without CMR as their standard practice. CCT was performed instead of CMR in patients with pacemakers or intracardiac leads (28/38, 74%). 88% of centers perform functional CCT, with 44% reporting ventricular volumes and function. Eight centers (19%) used a focused field of view on the RVOT, and branch pulmonary arteries, excluding the ventricles. RVOT measurement techniques varied: 84% of centers used angulated multiplanar reformats (MPR) and 28% used centerline technique/curved planar reformats (CPR). 88% used MPR-derived diameters and 24% used perimeter-derived area and average diameters. For CMR, only six centers (14%) reported using a non-gadolinium-based superparamagnetic iron oxide contrast agent. 30% of centers reported performing dedicated RVOT short-axis cine sequences and 16% reported utilizing 4D flow or 3D models. Median reporting times were 60minutes (IQR 60-90) for CCT and 120minutes (IQR 80-142) for CMR with dedicated technologist support available in 14 (33%) for CCT, and 15 (35%) for CMR. Overall, 84% favored a standardized imaging protocol and/or reporting template for PVR planning. CONCLUSION: This survey highlights wide variations in imaging practices for PVR planning, with most respondents supporting standardization, which may improve efficiency, quality, and potentially reduce radiation burden in this specific patient population.