4D flow CMR analysis in repaired tetralogy of Fallot: Where we are and where we are going

4D flow cardiovascular magnetic resonance (4D flow CMR) enable time-varying and multidirectional comprehensive retrospective analysis. However, this imaging modality must prove its relevance. After tetralogy of Fallot surgical repair, pulmonary valve function and right ventricle (RV) volumes are carefully followed. We investigate reliability of 4D flow CMR to measure these parameters compared to gold-standard 2D CMR. Twenty patients with rTOF and 5 healthy controls (mean age 29 ± 13 years, mean height 169 ± 7 cm, mean weight 66.2 ± 13.4 kg) underwent RV volumes and pulmonary regurgitation analysis by 2D CMR and 4D flow CMR using a clinical 1.5 T MRI scanner (GE; GE Healthcare, Milwaukee, WI, USA). Mean 4D flow acquisition time was 8.0 ± 1.8 min, whereas conventional 2D CMR was 39 ± 7.2 min ( P < 0.0001). Linear regression analysis shows a good correlation between 2D CMR and 4D flow CMR for RV stroke volume (SV) (r 2 = 0.89, P < 0.0001), end-diastolic RV volume (r 2 = 0.93, P < 0.0001), pulmonary regurgitation volume (r 2 = 0.94, P < 0.0001) and pulmonary regurgitation fraction (r 2 = 0.89, P < 0.0001). Mean bias [−1.96SD; +1.96SD] was 17.5 mL [4.75; 30.15] for RVSV, −8.4 mL [−17.04; 0.16] for end diastolic RV volume, 1.5 mL [−2.17; 5.23] for pulmonary regurgitation volume and 0.7% [−0.8; 2.2] for pulmonary regurgitation fraction by Bland–Altman analysis. Using 4D flow CMR, linear regression analysis showed a good correlation between RV SV and pulmonary valve systolic ejection volume (r 2 = 0.55, P < 0.0001), and between systemic and pulmonary flow (r 2 = 0.90, P < 0.0001). Mean bias was 22.7 mL [6.4; 39.0] for SV, and −0.06 mL [−0.18; 0.06] for systemic and pulmonary flow by Bland–Altman analysis. Vortex in the pulmonary artery trunk were observed in all cases of rTOF patients and limited analysis of pulmonary artery regurgitation flow. 4D flow imaging may be a clinically relevant alternative to 2D CMR in rTOF, providing faster and reliable data.