ABC | Volume 112, Nº4, April 2019

Original Article Ciuffo et al LA Remodeling and Dyssynchrony Arq Bras Cardiol. 2019; 112(4):441-450 analysis and Pearson’s correlation were also used to examine the relationship between LA intra-atrial dyssynchrony and LA-LGE. Four linear regression models are presented: Model 1 (unadjusted), Model 2 (adjusted for the following clinical characteristics: age, sex, type of AF, body mass index [BMI], history of heart failure, hypertension, and obstructive sleep apnea), and Model 3 (Model 2+ V min and S max ). Indices of LA intra-atrial dyssynchrony and LA-LGE were log-transformed due to non-normal distribution. We also evaluated the possibility of interaction between LA intra-atrial dyssynchrony and AF type. Pearson’s correlation coefficient was categorized with the following correlations: poor, 0; slight, 0.01-0.20; fair, 0.21‑0.40; moderate, 0.41-0.60; good, 0.61-0.80, and excellent, 0.81-1.00. In a subset of randomly selected participants (n = 15), a Bland-Altman analysis was performed to quantify intraobserver and interobserver reproducibility and inter-study reproducibility(21)(22) 23,24 . Moreover, the intraclass correlation coefficient (ICC) with a two-way randommodel was evaluated, in which agreement was categorized as follows: ICC, < 0.40, poor; ICC 0.40-0.75, fair to good; and ICC > 0.75, excellent. The statistical computations were performed using Stata, version 12.0 (StataCorp LLC, College Station, TX, USA). Results Clinical A total of 146 patients were included in the final analysis, and their clinical characteristics are summarized in Table 1. There were 61 (29.3%) female patients, and the mean age was 60.0 ± 10.0 years. A total of 102 patients (69.8%) had paroxysmal AF at the time of the procedure. Patients with paroxysmal and persistent AF were similar in terms of clinical baseline characteristics and medication usage, as demonstrated in Table 1; 4 of 44 patients (9.1%) in the persistent group and 2 of 102 patients (2.0%) in the paroxysmal group underwent cardioversion within 3-4 weeks prior to CMR (p = 0.158). Left atrial function, intra-atrial dyssynchrony, and atrial fibrillation type Patients with persistent AF had lower total LA emptying fraction (LAEF), active LAEF, SR, SR e , SR a , and left ventricular ejection fraction (LVEF) than those with paroxysmal AF (Table 2). In addition, SD-TPS was significantly higher in patients with persistent AF than in those with paroxysmal AF (median 3.6% versus 2.9 %, respectively, p = 0.036). SD‑TPS preA was not significantly different between the AF types (4.6% versus 3.7%, respectively, p = 0.227) (Table 2). The dyssynchrony analysis was performed in a consistent manner in all cases and took 5 ± 9 minutes per case. There was no difference in the amount of time required for the dyssynchrony analysis between the AF types (p = 0.35). LA Dyssynchrony and LA-LGE There was no significant difference in the extent of LA fibrosis quantified by LGE between the AF types (11.6 [6‑17.6]% of LA surface versus 13.8 [7.6-28.4] % of LA surface in the paroxysmal and persistent AF groups, respectively, p = 0.061). In Model 1, log-transformed SD-TPS and SD-TPS preA were associatedwith the LA degree of log-transformed LA-LGE enhancement (Table 3). After adjusting for age, sex, BMI, AF type, history of heart failure, Table 1 – Baseline characteristics Overall (n = 146) Paroxysmal AF (n = 102) Persistent AF (n = 44) p Clinical Age, years 60.0 ± 10.0 60.0 ± 10.1 59.7±9.8 0.906 Body mass index, kg/m 2 28.4 ± 5.5 28.0 ± 5.4 29.9 ± 5.3 0.073 Male, n (%) 102 (70.0) 74 (72.5) 28 (63.3) 0.134 Heart failure, n (%) 14 (9.6) 8 (7.8) 6 (13.6) 0.082 Coronary artery disease/vascular disease, n (%) 12 (8.2) 10 (9.8) 2 (4.5) 0.536 Diabetes, n (%) 15 (15.4) 12 (11.8) 3 (6.8) 0.704 Hypertension, n (%) 60 (41.1) 42 (41.2) 18 (40.9) 0.154 History of stroke/TIA, n (%) 9 (6.2) 8 (7.8) 1 (2.3) 0.351 CHA 2 DS 2 -VAS C 1.60 ± 1.5 1.5 ± 1.6 1.6 ± 1.2 0.942 Obstructive sleep apnea, n (%) 23 (15.8) 17 (16.7) 6 (13.6) 0.796 Ablation strategy (cryoablation), n (%) 34 (23.3) 28 (27.5) 6 (13.6) 0.324 Medication ACEI/ARBS, n (%) 37 (25.3) 24 (23.5) 13 (29.5) 0.389 Beta-blockers, n (%) 81 (56.3) 62 (60.8) 19 (43.2) 0.788 Calcium-channel blockers, n (%) 33 (22.9) 26 (25.5) 7 (15.9) 0.637 Number of antiarrhythmic drugs 1.2 ± 0.8 1.2 ± 0.8 1.4 ± 0.7 0.108 Data are presented as mean ± standard deviation, n (%), or median. AF: atrial fibrillation; TIA: transient ischemic attack; ACEI/ARB: angiotensin-converting enzyme inhibitor/angiotensin receptor blockers; CHA 2 DS 2 -VAS C : score for stroke risk assessment in atrial fibrillation. 444