Study population and exclusion criteria

Forty patients who had undergone first-time radiofrequency catheter ablation for long-standing persistent AF were enrolled from January 2009 and followed until May 2010 (ClinicalTrials.gov ID: NCT01144858 ). Long-standing persistent AF was defined as AF that has been present for 1 year or more, resistant to at least one electrical or pharmacological cardioversion and for which a rhythm control strategy was decided. Exclusion criteria included: age < 18 or > 80 years; severe valvular disease requiring surgery; valve prosthesis; known severe coronary artery disease; atrial and/or ventricular thrombosis; New York Heart Association functional class III–IV; cerebrovascular disease; pulmonary embolism; and latent or manifest hyperthyroidism. All patients gave their written informed consent.

Echocardiography study

A complete echocardiographic evaluation was carried out within 48 hours before the procedure using the conventional transthoracic approach and then transoesophageal echocardiography was carried out by two different physicians in a blinded fashion.

First, conventional transthoracic echocardiography was performed before and after AF catheter ablation (IE33 System; Philips Medical Systems, Andover, MA, USA) with routine echocardiographic measurements, using parasternal short and long axes, and apical four- and two-chamber views. The left atrium area was obtained via apical four-chamber zoomed views of the left atrium. Images and pulse Doppler flows of mitral inflow and tissue Doppler imaging at the mitral annulus were acquired from the four-chamber views.

Second, all patients were evaluated before catheter ablation by complete transoesophageal echocardiography with multiplane probes using a 7-MHz transducer (Vivid i; General Electric Medical Health, Horten, Norway). Left atrial spontaneous contrast and thrombus were sought. After a complete analysis of the left atrial appendage at the base of the heart with rotation of the probe between 0° and 180°, the incidence with the best alignment of the cursor with the appendage long axis was selected. The cursor was placed at the entry of the appendage for pulsed Doppler analysis and we considered the average value of 10 consecutive fibrillatory emptying waves ( Fig. 1 ) .

A. Pulsed Doppler tracing of left atrial appendage obtained by transoesophageal echocardiography showing left atrial appendage peak flow velocity (LAV) at the mouth of the appendage at 80° in a patient who achieved successful atrial fibrillation (AF) termination (0.42 m/s). B. Low LAV at the mouth of the appendage at 43° in a patient in whom persistent AF was not terminated during the procedure (0.22 m/s).

Outcome measures

The primary outcome was restoration of sinus rhythm during the catheter ablation procedure. The second outcome was absence of recurrence of atrial arrhythmias (AF, atrial tachycardia or flutter) during the year following the procedure.

Electrophysiological study and catheter ablation procedure

All patients received effective anticoagulation therapy (vitamin K antagonists, target international normalized ratio [IQR] of 2–3) for more than 1 month before ablation. This therapy was interrupted at least 48 hours before the procedure, with a heparin bridge. All antiarrhythmic drugs were discontinued 1 week before the procedure, except for amiodarone, which was maintained.

The electrophysiological study was performed under general anaesthesia using a standard protocol. The following catheters were introduced via the femoral vein: a steerable quadripolar catheter (Xtrem ^® ; Sorin Group, Le Plessis-Robinson, France) was positioned within the coronary sinus; a circumferential mapping catheter (Lasso; Biosense Webster, Diamond Bar, CA, USA) was introduced after transseptal access; and a 4-mm externally irrigated-tip ablation catheter (Thermocool, Biosense Webster) was used for mapping and ablation. After transseptal access, a single bolus of heparin (100 IU/kg body weight) was administered. The infusion was adjusted to maintain an activated coagulation time of 300 s or more. The transseptal sheath was also continuously infused with heparinized saline during the procedure.

Surface electrocardiograms (ECGs) and endocardial electrograms were continuously monitored and recorded for off-line analysis (Bard Electrophysiology, Lowell, MA, USA). Following transseptal catheterization, left atrial and coronary sinus electroanatomical mapping (Carto 3; Biosense Webster) was performed during spontaneous AF. Computed tomography registration and fusion of left atrial reconstructions with the electroanatomical map were subsequently performed. Endocardial AF cycle length was determined from intracardiac recordings at the left atrial appendage before ablation and was averaged for 30 consecutive intervals.

In all patients, sequential stepwise ablation described by O’Neill et al. , was performed by the same operator, blinded to the echocardiographic data. The procedure was terminated with the step that allowed AF conversion into sinus rhythm and no antiarrhythmic treatment was prescribed during the procedural period or during the procedure. In all cases, circular and linear lesions were verified after sinus rhythm restoration. The first step involved pulmonary vein isolation. The second step included linear ablation in the left atrium: a roof line was drawn between the right and left superior pulmonary veins and, if AF persisted, a mitral isthmus line was drawn from the mitral annulus to the left inferior pulmonary vein and coronary sinus defragmentation was performed. The third step consisted of electrogram-based ablation of complex fractionated atrial electrograms in the left and right atria . Complex fractionated atrial electrogram sites were tagged on the geometry obtained from three-dimensional mapping during AF. Lastly, a cavotricuspid isthmus line was performed only in patients with a history of common atrial flutter with ECG documentation.

When AF was converted to a regular arrhythmia, activation and entrainment mapping were performed to differentiate between focal and re-entrant mechanisms. Atrial tachycardias were targeted for ablation until sinus rhythm was achieved. When sinus rhythm was not restored by ablation, AF or atrial tachycardia was terminated by electrical cardioversion and the procedure was considered as a failure. After restoration of sinus rhythm, assessment of conduction block across the lines was performed in all patients . When necessary, supplemental radiofrequency energy was delivered to achieve block.

Discharge, follow-up plan and AF recurrence assessment

Treatment with vitamin K antagonists was resumed 1 day after the procedure and patients were discharged on day 3 receiving low-molecular-weight heparin until they had two consecutive international normalized ratios > 2. Patients were assessed before discharge and at the third, sixth and 12th months by clinical interview, echocardiography and 24-h Holter monitoring. In addition, patients were instructed to call their cardiologist in case of sustained palpitation, for immediate ECG recording. Vitamin K antagonists were prescribed for a minimum of 3 months and potentially discontinued in case of low thromboembolic score (CHADS ₂ score 0 or 1). Amiodarone was continued for at least 3 months in patients who were receiving amiodarone before the procedure and was interrupted in case of no recurrence at 3 months. Recurrence was defined as any symptomatic or asymptomatic atrial arrhythmia lasting > 30 s; it was evidenced by Holter monitoring at 3, 6 and 12 months or by 12-lead ECG in case of symptomatic palpitation at clinical interview.

Statistical analysis

Categorical variables are expressed as number and proportion, and continuous variables are expressed as mean ± standard deviation. Comparison of baseline characteristics of patients with and without AF termination by catheter ablation was performed using the χ ² test or Fisher’s exact test (as appropriate) for categorical variables and Wilcoxon’s test for continuous variables. All tests were two-sided. A P value < 0.05 was considered statistically significant.

LAV (highly associated with favourable outcome) was studied by analysis of receiver operating characteristics (ROC) to determine optimal cutoff values for the prediction of successful catheter ablation. ROC was evaluated using a plot of the true positive fraction (sensitivity) versus the true negative fraction (1–specificity) with a continuously varying decision threshold. The best cutoff value was defined as the point combining the highest sensitivity and specificity.

Age, duration of AF, LAV, left ventricular ejection fraction, left atrial area and AF cycle length from the left atrial appendage were considered in a logistic regression model to identify criteria associated with successful catheter ablation procedure. Categorized variables with a P value < 0.20 in univariate analysis were then considered in a logistic regression model to identify independent predictors of AF termination by ablation. A similar analysis was performed regarding the absence of AF recurrence during the 1-year period. All statistical analyses were performed using SAS software, version 9.1 (SAS Institute, Cary, NC, USA).