Assessment of the Dynamism of the Left Atrial Appendage Dimensions: A Computer Tomographic Analysis

Abstract

Background: Device sizing prior to left atrial appendage (LAA) closure is currently primarily based on transesophageal echocardiographic as well as invasive angiographic measurements, and can be challenging due to the complex and highly variable anatomy of the LAA. Computerized tomography (CT) is a 3-dimensional imaging modality that is increasingly being used for planning structural heart disease interventions. We assessed the variability of the measurements of the LAA ostium in patients with sinus rhythm and atrial fibrillation referred for CT angiography.

Methods: 101 consecutive patients with available retrospective spiral acquisitions as well as multiphase reconstructions between 0 to 90% of the peak R-wave to R-wave were included in this analysis. All acquisitions were performed using a third-generation dual-source system (Somatom Force, Siemens Healthineers, Forchheim, Germany). Data sets were transferred to dedicated Software (Ziostation2, Ziosoft Inc., Tokyo, Japan) which allows dynamic evaluation of the LAA ostium through the different phases of the cardiac cycle. Multiplanar reconstructions were aligned with the plane of the LAA ostium and measurements where performed in a cross-sectional plane orthogonal to the long axis of the LAA at the level of the left circumflex coronary artery. Four measurements were performed: area, circumference, area-derived diameter (√[area/Π] x 2) and circumference- derived (perimeter/Π). Furthermore assessment of the length if the LAA was assessed in all patients. Assessment of the Dynamism of the Left Atrial Appendage Dimensions: A Computer Tomographic Analysis Mohamed Marwan, MD*, Amina Vaillant, Fabian Ammon, MD, Daniel Bittner, MD, Michaela Hell, MD, Stephan Achenbach, MD Department of Cardiology, University of Erlangen-Nuremberg, Erlangen, Germany.

Results: Out of 101 patients (mean age 81 ± 8 years, 61% males), 48 patients were in sinus rhythm at time of acquisition and 53 patients were in atrial fibrillation. The mean area of the LAA ostium as well as perimeter were significantly larger in AF patients compared to SR patients (464±153 vs. 359±131 mm² and 78±12 mm vs. 69±12 mm for AF vs SR patients, respectively, p=0.001). Consequently the area derived diameter as well as perimeter derived diameter were consequently significantly larger in AF vs. SR patients (24±4 mm vs. 21±4 mm and 25±4 vs. 22±4 mm for area-derived vs. perimeter- derived diameter, respectively, p<0.001). The percentage difference between maximal and minimal LAA dimensions were significantly higher for sinus rhythm patients compared to atrial fibrillation [88% (IQR 60; 147%) vs. 21% (IQR 13; 42%), respectively, p<0.001] for median percentage area change and 34% vs. 10% for median percentage perimeter change (IQR 25; 52 vs. 7;18%, respectively, p<0.001). For atrial fibrillation patients, the largest LAA dimensions (area, perimeter, area-derived and perimeter-derived diameters) was measured at an average of 40% of the peak R-wave to R-wave whereas for sinus rhythm patients, the maximal LAA dimensions were measured at an average of 46% of the peak R-wave to R-wave (p>0.05). The mean length of the LAA was significantly larger in AF patients compared to SR patients (19.5 mm vs. 17 mm for AF vs SR patients, p=0.04) and the median percentage change in length was significantly higher in SR vs. AF (32% [IQR 19; 61%] vs. 13% [IQR 9; 19%] for SR vs. AF patients).

Conclusions: Dimensions of the left atrial appendage ostium vary significantly within different time points in cardiac cycle. These changes are more pronounced in patients in sinus rhythm compared to patients in atrial fibrillation which might impact sizing if CT is used for procedural planning prior to interventional closure of the LAA. According to our data, to identify maximal LAA dimensions, CT imaging for the purpose of LAA occlusion should be targeted in atrial diastole (40-50% of the peak R-wave to R-wave).