ABC | Volume 110, Nº4, April 2018

Original Article Muñoz et al Atrial fibrillation ablation in valve disease Arq Bras Cardiol. 2018; 110(4):312-320 Table 1 – Baseline characteristics of patients who maintained sinus rhythm in the immediate post-surgery period (3 months) (n = 44) Characteristics Age (years) 69 ± 9 Female gender, n (%) 32 (73%) Mitral surgery, n (%) 36 (82%) Aortic surgery, n (%) 16 (37%) Tricuspid intervention, n (%) 13 (29.5%) Cryoablation, n (%) 36 (82%) Antiarrhythmic treatment at discharge, n (%) 13 (29.5%) ACE inhibitors at discharge, n (%) 21 (48%) AF duration > 1 year before surgery, n (%) 26 (59%) LA biplane volume (mL/m 2 ) 68 ± 22 AP LA diameter (mm/m 2 ) 28.9 ± 5 LVEF (%) 63 ± 12 ACE: angiotensin converting enzyme;AF: atrial fibrillation;AP: anteroposterior; LA: left atrium; LVEF: left ventricular ejection fraction. In addition, 30 healthy individuals were analyzed following the same protocol in order to have a reference population for LA mechanics. Two experienced observers carried out the measurements in both populations in different times, in order to determine intra- and inter-observer variability. Intra-observer variability was calculated with measurements of the same rater in different moments including random samples of either patients or healthy controls. The same 2D echocardiographic loops of random samples of both patients and healthy controls were used for inter-observer variability. All images were digitally stored for offline analysis. Definition of AF recurrence AF recurrence was defined as presence of AF at any electrocardiogram or during at least 30 seconds in Holter monitoring. Statistical analyses Descriptive analyses were performed to explore study population characteristics. Categorical variables were reported as frequencies, and continuous variables with normal distribution were reported as mean ± SD. Median and interquartile range were used in cases of non-normality. Normal distribution of continuous variables was studied using Kolmogorv-Smirnov test. Differences among cohorts were analyzed using Chi-square test for categorical variables (or Fisher´s exact test when the comparison group was < 30 individuals), and Student- t -test (or Mann-Whitney test if the comparison group was < 30 and in case of non-normal distribution) for the numerical ones. Kaplan–Meier method was used for describing event free survival (AF) over time; the median was used as cutoff value to compare quantitative variables and differences between groups were investigated with the log-rank test. Those variables with p value < 0.15 were included for multivariate analyses using a Cox proportional hazard model. Significance was considered with a p value of < 0.05. Statistical analyses were performed using SPSS (Statistical Program for the Social Sciences [SPSS Inc., Chicago, USA]) version 15.0. Intra- and inter-observer agreements in the speckle- tracking measurements were studied by regression analyses and calculation of the intraclass correlation coefficient. Bland Altman plots, combined with calculation of 95% limits of agreement were also generated. For this analysis MedCalc Statistical Software version 15.6.1 (MedCalc Software bvba, Ostend, Belgium) was employed. Results AF recurrence was identified in 9 out of 53 cases in the immediate post-surgery period (3 months). These 9 patients were excluded for subsequent follow-up. The 44 patients with sustained SR after 3 months were included in the echocardiographic and rhythm follow-up. Baseline characteristics of this series are shown in Table 1. The majority of the population underwent mitral surgery (28 patients, 63.6%), 8 (18.2%) mitral and aortic, and only 8 required exclusively aortic intervention. Mitral valve surgery included 34 prosthetic replacement procedures (26 mechanical and 8 biological), and two mitral valve repair surgeries. Valve replacement was the procedure employed in all patients with aortic disease (12 mechanical and 4 biological). There were 13 (29.5%) tricuspid annuloplasties using Carpentier-Edwards ring in all cases. Overall, the study population showed preserved left ventricular ejection fraction (LVEF) and severely dilated LA. These patients were predominantly women with a mean age of 69 ± 9 years old (y.o.). Treatment at discharge included amiodarone in 30% of patients and angiotensin-converting- enzyme (ACE) inhibitors in 48% of patients. After a mean follow up of 17 ± 2 months, 13 new post‑operative cases of AF were identified. Myocardial deformation parameters (strain and strainR) for assessing LA mechanical function after SA were obtained from 1,245 left atrial segments that were correctly analyzed (71% of possible). On average, 15.5% and 19.4% of 24 potential segments were analyzed per patient and per control, respectively. LA mechanical function (strain and strainR) was significantly worse in all patients than in normal population, independently of SR maintenance (Table 2, Figure 2). As showed in Figure 3, intraclass correlation coefficient was always > 0.80, that represents good to excellent reliability and reproducibility of measurements. 13 The univariate analyses showed a trend of AF recurrence related to age, mitral surgery, cryoablation and LA biplane volume (Table 3). Patients with mitral valve intervention and cryoablation were younger (66.6 ± 8.4 vs 73.6 ± 9.1 y.o.; p = 0.041). As patients treated with cryoablation were the same as those with mitral intervention, cryoablation was 314