Concentration of apelin inversely correlates with atrial fibrillation burden

• Published in: Bratislavské lékarské listy Vol. 122; no. 3; pp. 165 - 171
• Main Authors: Bohm, A, Urban, L, Tothova, L, Bezak, B, Uher, T, Musil, P, Kyselovic, J, Lipton, J, Olejnik, P, Hatala, R
• Format: Journal Article
• Language: English
• Published: Slovakia 2021
• Subjects: Apelin; Atrial Fibrillation - diagnosis; Atrial Fibrillation - surgery; Catheter Ablation; Female; Humans; Pulmonary Veins; Recurrence; Treatment Outcome; atrial fibrillation; apelin biomarkers; catheter ablation
• ISSN: 0006-9248
• DOI: 10.4149/BLL_2021_026

Asymptomatic atrial fibrillation (AF) detection and pulmonary veins isolation (PVI) outcome prediction remain challenging. Our aim was to study the association between apelin and paroxysmal AF in patients undergoing radiofrequency catheter PVI. Sixty-three consecutive patients (55 ± 8years, 12 females) with paroxysmal AF without a structural heart disease and implanted ECG loop recorders undergoing PVI and healthy control group of 34 persons (41 ± 9.5years, 21 females) were included. Apelin plasmatic concentrations were measured before and three months after PVI. AF burden was continually assessed for three years. Apelin was significantly decreased in AF patients compared to the healthy controls (0.79 ± 0.09 vs 0.98 ± 0.06 ng/ml; p < 0.00001). Apelin plasmatic concentration of 0.89 ng/ml had 94 % specificity and 89 % sensitivity for AF prediction with the area under the curve (AUC) of 0.96. After propensity matching to sex, age and comorbidities, apelin concentration was significantly lower in AF group (0.78 ± 0.1 vs 0.99 ±0.06  ng/ml; p < 0.0001; AUC: 0.97). There was a significant inverse correlation between apelin concentration and AF burden both before and after PVI (Rho = ‒0.22; p = 0.05) and (Rho = ‒0.51; p = 0.006), respectively. There was no significant association between pre-PVI apelin and PVI long-term outcome. In patients without a structural heart disease apelin showed a significant specificity and sensitivity for AF prediction and inversely correlated with AF burden (Tab. 3, Fig. 3, Ref. 34).