PO-02-177 EPICARDIAL ADIPOSE TISSUE NEUTROPHIL INFLAMMATION RELATES TO SEVERITY BUT NOT TO RECURRENCE OF ATRIAL FIBRILLATION AND IS REFLECTED BY ATTENUATION IN CTA SCANS

Non-invasive tools for personalized, mechanism-based staging of the atrial fibrillation (AF) substrate are immature. Inflammation of the epicardial adipose tissue (EAT) is associated with increased atrial fibrosis in patients with AF. Inflammation correlates with attenuation on CT scans, interpreted as tissue density. However, how CT-EAT attenuation is associated with clinical AF severity, AF recurrence after ablation, and histological (local) inflammatory parameters, is largely unknown. To assess if (A) EAT CT scan attenuation and volume, and (B) EAT adipocyte size, infiltration of neutrophils, macrophages and lymfocytes, differ between patients with paroxysmal and persistent AF, and between those with and without recurrence after thoracoscopic AF ablation. All patients undergoing thoracoscopic AF ablation from 2017 to 2020 with a cardiac CT angiography scan before and at 6 months after surgery were included. Atrial and total EAT attenuation and volume were extracted (range -190 to -30 HU), and corrected for beam intensity and body surface area. Patients’ left atrial appendage (LAA) were excised and immunohistochemically stained for neutrophils (myeloperoxidase), macrophages (CD163, CD68), lymphocytes (CD3, CD4) and adipocytes (hematoxylin-eosin). All patients received standardized rhythm follow-up with 24h holters at 3,6, 9 and 12 months. In total, 134 patients were included, 27 had paroxysmal and 107 persistent AF, 44 were women, the CHA2DS2VASc score was 1.7 (±1.4), and 27 had a recurrence after 6 months of follow-up. From 115/134 patients a paraffinized LAA was available for immunohistochemistry. In persistent versus paroxysmal AF the CT-EAT attenuation of both atrial and total EAT was increased (-73.0±4.6 vs -75.3±5.3HU, -76.2±4.5 vs -78.4±5.1HU), while the EAT volumes were similar. Neutrophil count and adipocyte size correlated with EAT attenuation (r0.27, p0.005, r-0.21, p0.03, Spearman). Neutrophil count was higher in persistent vs paroxysmal AF (94.7 [188.7] vs 25.2 [68.5] cells/mm2, p<0.001) (figure). Patients with vs without AF recurrence had larger adipocytes 1780.0 [451.0], vs 1582.3 [398.3] μm2, p0.008, while they had similar EAT attenuation, EAT volume, and counts of the immune cells. CT-EAT attenuation reflects histological infiltration of neutrophils, and may be useful as a non-invasive marker to differentiate between AF types. CT-EAT attenuation does not predict AF recurrence after thoracoscopic ablation.