To investigate the accuracy of free-breathing CT in evaluating the volume and shape of epicardial adipose tissue (EAT), and further explore the characteristics of EAT volume and activity in patients with atrial fibrillation using ¹⁸F-FDG PET/CT " one-stop" imaging.

(1) Retrospective analysis was performed on 20 patients (16 males, 4 females, age: 33-86 (61.1±14.2) years) who underwent ¹⁸F-FDG PET/CT imaging and without obvious diseases affecting the images of the heart and surrounding lungs between March 2020 and May 2020 in Beijing Chaoyang Hospital. Free-breathing CT and breath-hold high resolution CT (HRCT) images were reviewed. Spearman rank correlation analysis, Bland-Altman consistency analysis and intraclass correlation coefficient (ICC) were used to evaluate the correlation and consistency of the EAT volume and shape, as well as the repeatability of the two operators′ measurements. (2) Prospective analysis was conducted to compare the differences in EAT volume and ¹⁸F-FDG uptake values between 20 patients (6 males, 14 females, age: 52-76 (66.0±6.4) years) with atrial fibrillation and 10 healthy controls (3 males, 7 females, age: 59-69 (66.0±3.6) years) collected between August 2017 and August 2018 in Beijing Chaoyang Hospital. Mann-Whitney U test was used to compare the differences in EAT volume and ¹⁸F-FDG SUV_max between patients with atrial fibrillation and healthy controls. EAT volume measurement was conducted by the combination of Mimics Research 21.0 software and manual analysis. The shape of EAT was automatically calculated by the same software to obtain the maximum length of the projection of the three-dimensional (3D) model on the reference axes (x, y, z). SUV_max of EAT was manually measured.

The measurements of EAT volume had good repeatability (intra-operator ICC=0.999; inter-operator ICC=0.997). There was a good correlation and a good consistency between EAT volumes measured by free-breathing CT and breath-hold HRCT (96.6 (79.9, 136.4) vs 96.2 (80.9, 135.8) ml; r_s=0.929, P<0.001); data of 19 cases were within 95% limits of agreement (95% LoA). The maximum projection length of EAT 3D model on the reference coordinate axis also showed good correlation and consistency (x axis: r_s=0.869, P<0.001, data of 19 cases were within 95% LoA; y axis: r_s=0.854, P<0.001, data of 18 cases were within 95% LoA; z axis: r_s=0.586, P=0.007, data of 20 cases were within 95% LoA). EAT volume of atrial fibrillation group was higher than those of healthy control group (137.2 (113.9, 202.9) vs 94.4 (76.6, 134.4) ml; z=-2.11, P=0.035) and SUV_max of EAT in the atrial fibrillation group was higher than that in healthy control group (1.2 (1.1, 1.5) vs 1.1 (1.0, 1.2); z=-2.14, P=0.035).

Free-breathing CT and breath-hold HRCT have good correlation, consistency and repeatability in measurement of EAT volume and shape. ¹⁸F-FDG PET/CT can be a " one-stop" imaging strategy for the evaluation of EAT volume and activity.

Adipose tissue; Positron-emission tomography; Tomography, X-ray computed; Fluorodeoxyglucose F18