Twenty-five patients with non-valvular atrial fibrillation and a high risk for stroke with a CHA₂DS₂-VASc Score of ≥ 1 admitted to our centre for percutaneous LAA closure between August of 2010 and August of 2015 were enrolled in this retrospective study. Nine patients were excluded due to lack of adequate echocardiographic evaluation before or after the procedure or poor echocardiographic window for the evaluation of left atrial mechanics.

Referral indications for percutaneous LAA closure were contra-indication for long-term oral anticoagulation, bleeding events during oral anticoagulation, labile international normalized ratio (INR) or embolic events despite proper anticoagulation.

Clinical data included past medical history, current medication, the CHA₂DS₂-VASc and HAS-BLED scores, and diagnostic evaluation by routine laboratory testing, electrocardiography, and echocardiography.

Sixteen patients with good echocardiographic window for assessment of the left atrial mechanics were included in our study.

The study was approved by the ethics committee of our institution.

Echocardiography was performed on the day before and 3 months after percutaneous closure of the LAA.

Echocardiographic examinations were performed using an ultrasound system (Vivid 7, General Electric^®, Horten, Norway) and tissue harmonic imaging at 1.7/3.4 MHz. A complete echocardiographic study was performed using standard views according to current guidelines.14 Three consecutive heart cycles were acquired for quantification of the left atrial size and 2D-STE analysis for sinus rhythm patients, and five consecutive heart cycles were obtained for atrial fibrillation patients.

Left atrial volume was assessed by the biplane method of disks from the apical 4- and 2-chamber views and the measurements were indexed to the body surface area according to established recommendations.15 Minimum left atrial volume was measured at left ventricular end-diastolic volume, and maximum left atrial volume at end-systole. Left atrial emptying fraction was calculated as (maximum left atrial volume - maximum left atrial volume)/ maximum left atrial volume.15

The 2D-STE method was used to calculate regional and global longitudinal ƐR and SR_R (Figure 1). A minimum frame rate of 60 frames/sec was required for a reliable operation of the program. The recordings were processed using an acoustic-tracking dedicated software (EchoPAQ 9.0, GE Healthcare^®, Horten, Norway), which allowed for an off-line semi-automated analysis of speckle-based strain. Left atrial endocardial surface was manually traced in end-systole in both four- and two-chamber views by a point-and-click approach. An epicardial surface tracing was then automatically generated by the system, generating the region of interest (ROI). For definition of the ROI at the discontinuity of the left atrial wall (corresponding to pulmonary veins and left atrial appendage), the limit of left atrial endocardial and epicardial surfaces at the junction of these structures was extrapolated. After manual adjustment of ROI width and shape to ensure optimal tracking, the software divided the ROI into six segments (basal, middle and apical segments of the atrial septum and lateral wall), and the tracking quality of each segment was automatically scored as either acceptable or non-acceptable, with possible further manual correction. Segments from which good quality images could not be obtained were rejected by the software and excluded from the analysis. In subjects with good quality images, a total of twelve segments were analyzed. The software displayed peak longitudinal ƐR and strain rate for each of the twelve segments and the average global strain. Peak ƐR were expressed in percentages and SR_R in s^-1. Since left atrial wall strain is reliably imaged and is not constrained by other cardiac chambers, recent consensus of imaging for evaluation of atrial fibrillation patients recommend the evaluation of this parameter rather than global ƐR.16 Therefore, we also performed a comparison between left atrial lateral wall strain and SR_R at baseline and after device implantation. Since we included patients with atrial fibrillation and sinus rhythm, we used the first left ventricular systolic frame as the frame of interest - QRS timed analysis.

Figure 1

LAA closure device was implanted in the catheterization laboratory. The device used was an Amplatzer^® (St. Jude Medical, St. Paul, Minnesota, USA) and was delivered through an appropriate sheath depending on the size of the selected occluder through a puncture in the femoral vein. Deployment and position of the device were controlled by fluoroscopy, and by periprocedural transesophageal or intracardiac echocardiography. LAA was reached through a transseptal puncture. Decision on device size was made upon anatomical morphology, and measurements in echocardiography and fluoroscopy. Oral anticoagulation, if present, was discontinued 48 hours prior to the procedure. During procedure, heparin was administered with an activated clotting time of 250s. Dual antiplatelet therapy with aspirin 100 mg and clopidogrel 75 mg was recommended for 1 month, followed by long-term antiplatelet therapy with aspirin 100 mg daily. No oral anticoagulation was recommended after device implantation.

The Kolmogorov-Smirnov test was used to evaluate the distribution of the continuous variables. In the overall sample, all variables were normally distributed, except for ƐR and follow-up time; and when patients were separated by group (no change and decrease of ƐR and SR_R), the variables were not normally distributed. According to distribution normality, continuous data were presented as mean and standard deviation or as median and interquartile range. Quantitative variables with normal distribution were compared by the t-test and quantitative variables without normal distribution by the Mann-Whitney test. Qualitative variables were compared using the chi-square test. Differences between baseline and post-implantation of the LAA occlusion device were analysed by the paired sample t-test.

Statistical analysis was carried out with SPSS^®15 and GraphPad Prism^® 6.05. A two-tailed p value < 0.05 was considered statistically significant.

Mean age of our sample was 71 ± 9 years, with male predominance (63%). Seventy-five percent of patients had permanent atrial fibrillation. There was no history of percutaneous atrial fibrillation ablation attempt or surgical Maze procedure. Our population had a high embolic and bleeding risk, expressed by a median CHA₂DS₂-VASc score of 5 [4-5] and HAS-BLED score of 3 [2-3].

Major bleeding (62%) was the most common indication for the procedure, followed by labile INR (19%), embolic events despite anticoagulation (13%), and poor compliance with anticoagulation medication (6%). Percutaneous LAA closure was performed successfully in all patients using the cardiac plug device (size, 24 ± 2 mm), without any major complications during or after the procedure.

Characteristics of the study population are summarized in Table 1.

Maximum and minimum values of left atrial volume and the left atrial emptying fraction before and after the procedure are represented in Fig. 2 and Fig. 3, respectively. No differences were found in maximum left atrial volume (44 ± 11 vs. 46 ± 13 mL/m²; p = 0.54), minimum left atrial volume (32 ± 8 vs. 37 ± 14 mL/m²; p = 0.09), or the left atrial emptying fraction (26 ± 17% vs. 21 ± 14%; p = 0.33) after the intervention compared with the baseline values.

Figure 2

Figure 3

Global and regional peak ƐR and SR_R of the 12 segments before and months after percutaneous closure of the LAA are listed in Table 2. Similar values of ƐR (10.1 [8.1-14.7] vs. 12.7 [5.4-16.5]%; p = 0.81) and SR_R (1.06 ± 0.26 vs. 1.13 ± 0.34 s^-1; p = 0.38) were observed before and after the procedure (Figure 4).

Figure 4

Assessment of left atrial lateral wall revealed similar ƐR (11.0 [6.5-19.8]% vs. 8.2 [2.7-15.9]%; p = 0.60) and SR_R (1.01 [0.78-1.54] vs. 1.02 [0.85-1.56] s^-1; p = 0.75) before and after the procedure.

In 44% of patients, there was a decrease in ƐR. There were no differences regarding patient age, baseline left atrial volume, left atrial emptying fraction, ƐR, SR_R, CHA₂DS₂-VASc or HAS-BLED scores, cardiac-plug device size, or incidence of cardiovascular adverse events during follow-up between patients with decreased and unaltered postoperative ƐR values (Table 3).

Our study has several limitations. First, this was a single-center study with a relatively small sample size. However, there is a paucity of data regarding the effects of percutaneous LAA closure on left atrial function in the literature. Second, the retrospective nature of the study limited the evaluation of additional clinical and analytical parameters. Third, although we analyzed the impact of LAA percutaneous closure on left atrial mechanics, the design of our study made the assessment of clinical outcomes impossible. Finally, most of our patients had atrial fibrillation and left atrial dysfunction at baseline. It would therefore be important to study the influence of LAA percutaneous closure on left atrial function in patients who are in sinus rhythm with normal or slightly altered left atrial function.

Further studies including large populations of patients in sinus rhythm and in atrial fibrillation are needed to provide definitive evidence of the impact of LAA occlusion not only on left atrial physiology at long term, assessed by 2D-STE, but also on clinical outcomes.