Susceptibility Artifact in Wideband Late Gadolinium Enhancement Cardiac Magnetic Resonance Imaging of Patients with Cardiac Implantable Devices (RF_FR_304)

Friday, January 31, 2025

9:40 AM – 9:50 AM East Coast USA Time

Location: Blue Room PreFunction

Presenting Author(s)

TL

Thippayaporn Lopaisankrit, MD

Radiology Fellow (Body Imaging)
King Chulalongkorn Memorial Hospital, Thailand

Primary Author(s)

TL

Thippayaporn Lopaisankrit, MD

Radiology Fellow (Body Imaging)
King Chulalongkorn Memorial Hospital, Thailand

Co-Author(s)

YV

Yongkasem Vorasettakarnkij, Co-first author

Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thailand
MT

Monravee Tumkosit, MD

Assistant Professor
Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thailand

Background: The metallic components of cardiac implantable electronic devices (CIEDs) can cause susceptibility artifacts (SA) on cardiac magnetic resonance imaging (CMR), potentially interfering with image interpretation. Previous studies have suggested an association between certain parameters measured on chest radiography (CXR) and the occurrence of these artifacts. However, these studies were conducted prior to the introduction of wideband late gadolinium enhancement (WB-LGE) techniques, which significantly reduce SA. This study aims to assess the prevalence of SA in patients with CIEDs using WB-LGE and to identify CXR parameters that associated with the occurrence of these artifacts.

Methods:

A prospective study was conducted involving patients with CIEDs who underwent 1.5 Tesla CMR from March 2022 to July 2024. WB-LGE images were acquired 5-10 minutes after gadolinium administration, and all 17 segments of the left ventricle (LV) were assessed for CIED-related SA. SA was classified into grade 0 (no artifact), grade 1 (artifact affecting < 50% of myocardial thickness), and grade 2 (artifact affecting ≥ 50% of myocardial thickness or precluding scar evaluation). CXR was evaluated for pulse generator (PG) volume, lead dimensions, and various distances between the PG and cardiac borders. A multiple logistic regression model was used to identify factors associated with SA.

Results: Forty patients were included (mean age 61 ± 16 years, 68% male). SA was observed in 70% of patients, with the highest prevalence among those with cardiac resynchronization therapy with defibrillators (100%) and implantable cardioverter defibrillators (89%) compared to permanent pacemakers (3%). The most commonly involved LV segments were the basal anterior (70%), mid anterior (65%), and apical anterior segments (55.0%). SA was significantly more common in patients with younger age (57 ± 14 vs 70 ± 16 years), shockable devices (90% vs 18%), larger PG volume (30 ± 6 vs 17 ± 9 cm³), and right ventricular (RV) lead diameter (7.8 ± 1.0 vs 6.2 ± 1.4 Fr). The nearest distances from the PG inferior border to the cardiac border, the intersection point of the cardiac apex and the left diaphragm, and the most lateral left cardiac border were significantly smaller in patients with SA, while the other distances were not. On multivariate analysis, a larger RV lead diameter (adjusted odd ratios (aOR) 3.0, 95% CI 1.3-7.0; p = 0.009) and a smaller distance from the PG inferior border to the most lateral left cardiac border (aOR 0.72, 95% CI 0.53-0.98; p = 0.037) were independent predictors of SA. Optimal cut-offs were determined for RV lead diameter (≥ 6.8 Fr) and the PG-to-lateral border distance (< 10.8 cm).

Conclusion: Despite the use of WB-LGE, SA remains prevalent in CMR of patients with CIEDs, particularly in those with defibrillators. Larger RV lead diameter and proximity of the PG to the cardiac border are independent predictors of SA.