Percutaneous closure of the left atrial appendage (LAA) facilitates stroke prevention in patients with atrial fibrillation. Thereby, accurate LAA assessment is required for optimal device selection. However, underestimation of the LAA diameter is observed by utilizing transesophageal echocardiography (TEE) and x-ray fluoroscopy (XR), the current gold standard imaging techniques for LAAc. We investigated the application of non-contrast-enhanced cardiac magnetic resonance imaging (cMRI) to support preprocedural planning for LAA closure (LAAc) as a non-invasive, radiation-, and contrast agent-free 3D imaging modality.
Methods
13 patients (85% male, 76±9 years) with atrial fibrillation (paroxysmal, persistent, long-persistent, permanent) and bleeding during oral anticoagulation undergoing LAAc were involved. Based on preprocedural cMRI, the maximum, perimeter- and area-derived diameters of the LAA were quantified and compared to the XR- and 2D TEE values derived periprocedurally (Fig.1). Further, optimal C-arm angulations with the projection direction perpendicular to the landing zone determined based on cMRI (Fig.2) were compared to the angulations used for periprocedural XR measurement.
Results
Regardless arrhythmia, the LAA could be identified and measured on the non-contrast enhanced cMRI in all patients. Diameters derived from perimeter and area based on cMRI showed great congruency compared to those measured by XR, whereas the maximum diameter exceeded (Tab.1). Compared to TEE assessment, cMRI-derived diameters were significantly larger. In about 40% of the cases it was necessary to additionally acquire contrast-enhanced XR in an angulation deviating from the classical projection range for LAAc, what was predicted based on cMRI.
Conclusion
The potential of non-contrast-enhanced cMRI to support the preprocedural planning of LAAc was demonstrated. Regardless of atrial fibrillation, based on cMRI the shape of the LAA could be assessed in 3D and a landing zone with respective dimensions and optimal XR angulations could be identified. Diameter measurements based on LAA area and perimeter well correlated with the actual device selection parameters. The angulation prediction based on cMRI showed high potential of saving radiation, contrast agent, and interventional time.
Disclosure
The project on which this abstract is based was funded by the Federal Ministry of Education and Research (code: 13GW0372C). Responsibility for the content lies with the authors.
Fig.1 Exemplified diameter assessment: The diameter of the landing zone measured in TEE (A), XR (B) and cMRI (C,D). Based on cMRI, maximum (C, white solid line), minimum (C, orange dashed line), perimeter-derived (D, white dashed line), and area-derived (D, orange area) diameters are acquired.
Fig.2 Angiographical simulations with ostium (red) and landing zone (green) of the LAA determined based on cMRI in classical angulation range (A) and patient specific optimal angulation (B).
Tab.1 Comparison of LAA diameters derived from XR, TEE, and cMRI (maximum: dmax; perimeter-derived: dperi; area-derived: darea).
| Mean deviation ± standard deviation in mm; P-value | Pearson correlation coefficient; P-value | |
| XR vs TEE | 1.3 ± 2.2; 0.09 | 0.85; <0.05 |
| dmax vs TEE | 4.6 ± 2.9; <0.05 | 0.77; <0.05 |
| dperi vs TEE | 2.2 ± 2.0; <0.05 | 0.87; <0.05 |
| darea vs TEE | 1.6 ± 1.7; <0.05 | 0.90; <0.05 |
| dmax vs XR | 3.3 ± 3.5; <0.05 | 0.68; <0.05 |
| dperi vs XR | 0.9 ± 2.8; 0.32 | 0.77; <0.05 |
| darea vs XR | 0.3 ± 2.6; 0.70 | 0.79; <0.05 |
https://dgk.org/kongress_programme/jt2023/aV124.html