Introduction

Four-dimensional flow cardiovascular magnetic resonance (4D flow CMR) allows quantification of biventricular blood by flow components and kinetic energy (KE) analyses generating unique insights for comprehensive hemodynamic assessment. This study aimed to compare the differences of 4D flow parameters for left ventricle (LV) and right ventricle (RV) in repaired tetralogy of Fallot (rTOF) and age-matched controls, and identify sensitive markers to differentiate rTOF with preserved RV ejection fraction (EF) from controls.

Methods

54 rTOF patients (14 pediatric, 40 adult; age 29 ± 16 years; 26M:28F) and 54 age-matched controls (14 pediatric, 40 adult; age 31 ± 16 years; 27M:27F) were recruited from four centers. CMR acquisition was performed using 3.0T Ingenia (Philips Healthcare, he Netherlands), 1.5T Magneton Tim Trio (Siemens Medical Solutions, Erlangen, Germany), and 3.0T (GE Healthcare, Waukesha, WI) magnetic resonance scanners. Typical 4D flow scanning parameters are summarized in Figure 1. Volumetric, 2D phase-contrast flow and 4D flow analyses were performed using MASS software (Leiden University Medical Center, Leiden, The Netherlands). Individual pathlines were generated from velocity vectors for all voxels found within the segmented LV or RV endocardial borders. Depending on pathline trajectories within the LV or RV during the analyzed cardiac cycle, four flow components were discernible: (1) direct flow enters and exits the chamber in the same cycle; (2) retained inflow enters but does not exit the chamber during the cycle; (3) delayed ejection flow starts within the chamber and exits during the cycle; and (4) residual volume remains in the chamber for at least two cycles. LV and RV flow components were indexed to LV or RV end-diastolic volumes (EDV), respectively. For each voxel, kinetic energy (KE) was calculated as KE=1/2·ρ_blood·V_voxel·v², where ρ_blood is blood density (1.06 g/cm³); V_voxel, voxel volume; and v, velocity magnitude. By summing individual voxel KE values across all time points, the total KE throughout the cardiac cycle is obtained. All KE parameters are normalized to EDV (KEi_EDV) with units expressed in μJ/ml. KEi_EDV parameters - peak systole, average systole and peak E-wave - were extracted from the time-resolved KE curve (Figure 2). In the RV, we defined efficiency index = effective cardiac index/RV systolic KEi_EDV, where effective cardiac index = effective cardiac output/body surface area and effective cardiac output = heart rate x (RV stroke volume – pulmonary backward flow). rTOF patients were stratified into preserved RV ejection fraction (EF) ≥45% and reduced RVEF <45% groups for comparison. Statistical analysis was performed using SPSS 22, and P value <0.05 was considered significant.

Results

4D flow analysis was feasible in all subjects. See Figure 3 and Figure 4 for results. rTOF patients had significantly greater RV volumes and lower RVEF, effective cardiac output and effective cardiac index compared with controls. rTOF patients had significantly lower RV direct flow and efficiency index, higher RV residual volume, peak systolic, systolic, and peak E-wave KEi_EDV, lower LV delayed ejection flow, higher LV retained inflow and peak E-wave KEi_EDV than controls (all P<0.05). RV direct flow progressively decreased and RV residual volume progressive increased from controls to rTOF with preserved EF to rTOF with reduced EF (all P<0.05). ROC analysis showed that RV direct flow and efficiency index were more sensitive (AUC=0.805 and 0.806) for discriminating rTOF with preserved EF from controls (Figure 5).

Conclusion

RV kinetic energy and flow components analysis from 4D flow CMR in rTOF patients showed that reduced efficiency index and RV direct flow were superior to traditional measurements for discriminating rTOF with preserved RVEF from controls.

Acknowledgements

No acknowledgement found.

References

No reference found.