|| Checking for direct PDF access through Ovid
PURPOSE: Device closure of the left atrial appendage (LAA) is an emerging therapy for patients at high risk of thromboembolic stroke unable to be safely anticoagulated. Complete occlusion or coverage of the appendage is required and two currently available devices (Watchman, Atritech Inc, Plymouth MN / Amplatzer Cardiac Plug, AGA Medical Corp, Golden Valley MN) are usually sized according to 2D transoesophageal echocardiography (TOE). Ex-vivo descriptions of LAA specimen have emphasised its oval orifice structure. Using 2D TOE may significantly underestimate size leading to undesired LAA flow after device occlusion. We hypothesised that more accurate LAA orifice measurements during device closures can be achieved by 3D TOE. METHOD: We acquired standard 2D LAA views from the mid-oesophageal 90° position and 3D zooms using Philips iE33 platform with a X7-2t matrix array in 41 patients during routine TOE. The images were taken by aligning a biplane region of interest over the LAA ensuring the whole of the LAA orifice, left upper pulmonary vein and lateral portion of the mitral annulus were included. 3D volumetric data were processed in Qlab7.1. Multiplanar reconstruction (MPR) was used to define en-face views of the LAA orifice. The longest diameter was measured and the short diameter taken at 90° to this. The area of the orifice was measured by planimetry, and the non-circular shape expressed as a ratio of the longest diameter divided by the orthogonal diameter (a circle would = 1). LAA working depth was defined as a perpendicular distance from the orifice to the LAA back wall. Image acquisition, data storage and analysis were performed in accordance with departmental protocols. We assessed differences in means (standard deviation, range) using an unpaired t-test. RESULTS: 2D analysis gave an LAA orifice diameter of 17.4mm (4.3mm,10–29mm) whereas MPR gave a diameter of 26.4mm (6.6mm,12–39.4mm), p<0.001. The working LAA depth was 17.6mm (5.2mm,7.8–28mm) on 2D and 18.8mm (6.0mm,7–34.4mm) on MPR (p 0.336). Simulation of the traditional 2D 90° view by MPR yielded a diameter of 15.9mm (4.0mm,9.2–26mm), p0.11. Therefore, the use of MPR itself does not overestimate LAA diameter. All LAA orifices were oval with a diameter ratio of 1.4 (0.25,1.1–2.0). CONCLUSION: 2D TOE of the LAA orifice diameter significantly underestimates size by a mean of 9mm (max by 17.4mm). 3D MPR allows identification of the true oval shaped orifice. This may aid current LAA occlusion device selection. 3D TOE visualisation of in-vivo LAA anatomy could influence new non-circular device designs leading to improved LAA occlusion results.