108/Relationship between left atrial wall thickness left atrial morphology and stroke history in the SCOT-HEART trial

Background: Detection of atrial fibrillation after stroke may be challenging and often requires prolonged monitoring. Atrial fibrillation is known to be associated with imaging markers of atrial structural remodelling such as left atrial dilatation and fibrosis. However, assessment of atrial morphology is challenging with previous approaches lacking standardisation.
Objective: To determine if left atrial morphological parameters assessed using computed tomography with a standardised semi-automatic algorithm are associated with a history of stroke and could be a novel tool to guide therapeutic decisions in patients following stroke.
Methods: Data from the SCOT-HEART trial cohort were retrospectively analysed including 71 patients with a history of stroke, 71 control patients with no stroke or atrial fibrillation and 27 patients with atrial fibrillation. A systematic method of segmentation and morphological assessment was used to derive left atrial volume, shape, wall thickness, mass, sphericity and appendage morphology.
Results: There were no differences in median left atrial volume, left atrial appendage volume or left atrial sphericity between the three groups. Median wall thickness in the Stroke Group, Control Group and AF Group were 1.71 ± 0.51 mm, 1.54 ± 0.45 mm and 1.69 ± 0.55 mm, respectively (p=0.184). There were no differences in the frequencies of different appendage morphologies between groups (p=0.191). On multivariate analysis, current smoking status and CHA2DS2VaSc score were predictive of stroke history but the addition of atrial morphological parameters had little incremental impact on the predictive model for stroke.
Conclusions: Patients with stroke or atrial fibrillation have no major differences in atrial morphological parameters suggesting that these characteristics are not major determinants of clinical disease.