Introduction

3D non-contrast-enhanced MR angiography (MRA) has been established for morphological assessment of congenital heart disease, whether with bSSFP-based readout at 1.5T or Dixon at 3.0T^1,2. However, signal loss due to off-resonance effects at tissue-air-interfaces in the lung precludes evaluation of pulmonary vessels, particularly in children at high field. The purpose of this study was to investigate an optimized T2prep pulse with 16 composite refocusing pulses (MLEV16) for improved PV visualization in pediatric patients with congenital heart disease (CHD) at 3.0T, with comparison to the standard T2prep pulse using MLEV4 and/or contrast-enhanced MRA.

Methods

Previously introduced Relaxation-Enhanced Angiography without ContrasT (REACT)^3-5, based on magnetization preparation pulses of non-selective T2-prep and inversion recovery and data acquisition using 3D dual-echo turbo-field echo (TFE) Dixon with semi-flexible echo times, was applied for non-CE thoracic MRA at 3.0T. Standard T2prep (4 composite refocusing pulses “MLEV4”, TE = 50 ms, interval = 12.5 ms), and TR/TE1/TE2 = 4.8/1.47/2.9 ms) was used as reference (REACT-Standard). In addition, an optimized T2prep with shorter refocusing interval (“MLEV16”, TE = 50 to 60 ms, 16 refocusing pulses, interval = 3.125 to 3.75 ms)⁶ that is less sensitive to ΔB0 and flow was implemented (REACT-MLEV16) and compared to classic REACT as an alternate approach to minimize signal loss^7,8. Figure 1 shows the pulse sequence diagram. Other imaging parameters were: FOV = 300 × 300 mm², Matrix = 200 × 200, acquired voxel size = 1.5 × 1.5 × 1.5 mm³, inversion delay = 60 to 70 ms, flip angle = 12˚, nominal scan time = approximately 3:15 min). Compressed sensing reconstruction in combination of wavelet transformation and sensitivity encoding (SENSE) coil information (compressed SENSE) was applied with an acceleration factor of 6 for scan reduction. Two healthy young adult volunteers and 16 patients aged under 18 years regardless of the type of CHD or previously performed surgical procedures were scanned on a 3.0T thoracic clinical MRI system (Philips Ingenia Elition, Best, the Netherlands). Steady-state contrast-enhanced MRA was performed in 5 patients during administration of gadobutrol (0.1 mmol/kg body weight) with slow flow rate. All MRA scans, whether native using REACT-Standard and REACT-MLEV16 or CE-steady state, were respiratory navigator gated and ECG triggered to end-diastolic single cardiac phase.
The proposed pre-pulse modifications increase SAR on 3T but fall within acceptable limits. Average SAR for standard MLEV4 and MLEV16 were below 2.0 W/kG.
PV and other major vessels were visually inspected and evaluated. Because in most patients the pulmonary vessels such as RPA suffer from severe signal suppression related to ΔB0, a proper signal ratio measurement (e.g., RPA/RV or PV/LV) was not possible. The myocardium/RV ratio was analyzed as a measure of myocardial suppression.

Results and Discussion

Overall image quality was found better for REACT-MLEV16 compared to REACT with standard T2prep pulse in terms of signal strength, sharpness, and artifacts. REACT-MLEV16 provided visually much better delineation of pulmonary veins and pulmonary arteries. One representative case example of both standard MLEV4 and MLEV16 T2prep is shown in Figure 2 in one pediatric patient. The attempt to reduce the ΔB0 and flow sensitivity provided much brighter and homogeneous signal in all three orthogonal orientations from the originally acquired isotropic dataset. It is to be noted that visualization and delineation of pulmonary vessels using non-contrast-enhanced MRA at 3.0T, whether balanced (bSSFP) or non-balanced signal readout, in children is very often not possible due to complex e.g. turbulent flow induced by, for example, severe pulmonary insufficiency during diastolic phase. This made a measure of the signal ratio across the vessels extremely difficult. Image quality and visualization of the pulmonary vessels were considered comparable between REACT-MLEV16 and steady-state MRA steady-state-MRA. Figure 3 demonstrates the comparison in one clinical case in a young child after mitral valve replacement. Myocardium-to-blood contrast ratio was comparable between REACT-MLEV16 and steady-state-MRA, same applied to vessel diameter measurement (not reported here due to low case number and statistical power at this stage). A more detailed study to compare the effect of blood flow complexities and local field inhomogeneities in signal suppression in PV as previously implied⁶ needs to be further investigated.

Conclusion

Non-contrast thoracic or extracardiac MRA using optimized T2prep based on MLEV16 pre-pulse can improve visualization of the pulmonary vessels at 3.0T, even in children with challenging vascular and flow abnormalities. It allows better clinical evaluation of congenital heart disease in comparison to standard T2prep with MLEV4 pre-pulse. Though initial clinical results are promising, further studies are needed to investigate its clinical performance in larger cohorts.

Acknowledgements

The authors thank Joshua S Greer (Philips USA) for valuable discussion.