Late Gadolinium Enhancement with Fat-Water Separation in Patients with MRI-Conditional Cardiac Implantable Electronic Devices at 3T (RF_SA_454)

There is a growing need for CMR imaging of patients with cardiac implantable electronic devices (CIED), particularly for assessing scars with late gadolinium enhancement (LGE).¹ Although many recent CIEDs are approved for imaging at 3T, CMR imaging for these patients is rarely performed at 3T. Recent data shows that 2/3 of CMR studies in heart failure patients from the SCMR registry were imaged at 3T,² highlighting the growing need for robust imaging of CIED patients at 3T. In this study, we sought to develop an LGE imaging sequence using a spoiled GRE readout with two echoes, allowing fat-water separation using the Dixon technique to reduce artifacts related to poor fat suppression at 3T in patients with CIED.

Methods:

A 2D ECG-triggered GRE research sequence was implemented with two bipolar echoes and a non-selective adiabatic hyperbolic secant inversion recovery pulse with a spectral bandwidth of 4.0 kHz (Figure 1A). Imaging parameters were as follows: TE₁/TE₂ 1.21/2.64 ms, FA 12°, readout FOV 380 mm, in-plane resolution 1.6 mm, slice thickness 8 mm, bandwidth 801 Hz/px, GRAPPA R=2. Imaging was performed every other heartbeat to allow for signal recovery. Water and fat images were reconstructed using the two-point Dixon algorithm.³

All imaging was performed at 3T (MAGNETOM Vida, Siemens Healthineers, Forchheim, Germany). A phantom comprising doped agarose vials and two oil bottles was scanned with applied B₀ field variation to test the effectiveness of fat-water separation in the presence of B₀ inhomogeneity. Eight patients (7 males, 66 ± 14 years) with implanted cardiac defibrillators (ICD, N=6) or pacemakers (N=2) were scanned following informed consent. Patients were positioned with arms above their heads and imaged during end-inspiration to maximize the distance between the device and the region of interest. LGE imaging was performed 15-20 minutes after injection of 0.1 mL/kg Gadovist (Bayer, Berlin, Germany). Optimal inversion times (TI) were determined using a free-breathing Look-Locker sequence.
 


Results:

Water-only and fat-only images obtained in a phantom with applied B₀ field variations are shown in Figure 1B. Dixon fat-water separation is robust in the presence of a large applied magnetic field gradient, and the vial representing a healthy myocardium is successfully nulled. Fat-water separation in the heart was successful in all subjects. Figure 2 shows example water-only and fat-only images in multiple cardiac views in a patient with ICD, where robust water-fat separation is achieved. Examples of myocardial scar detection in patients with ICDs are shown in Figure 3. On in-phase images, it is challenging to differentiate scar tissue from fat, which also appears bright. Regions of LGE are visible in the water-only images.
 


Conclusion:

LGE-Dixon enables robust fat suppression for LGE imaging in patients with CIED at 3T. Future studies will compare image quality with conventional methods in a larger patient population.