Methods
All patients undergoing CA for AF at our institution between January 2006 and August 2018 were analyzed including repeated procedures for AF and/or left-sided atrial tachycardias. All patients signed informed consent about the procedure. The study was approved by the institutional review board and conducted according to the Helsinki declaration. The need for informed consent was waived for the study.
Ablation procedure
Bilateral femoral veins were punctured either using anatomical landmarks or under direct ultrasound (US) guidance (from 2016). In the most common setting, two Swartz SL1 8.5 Fr sheaths were introduced via the right femoral vein and two short sheaths via the left femoral vein: 7 Fr for coronary sinus catheter and 11 Fr for intracardiac echocardiography probe. Between 2006 and 2012, the right jugular / subclavian access was used for coronary sinus catheter placement. For robotic cases (Sensei, Hansen Medical, Mountain View, CA, USA), which were performed between 2008 and 2011, 14 Fr sheath was employed to accommodate Artisan electromechanical catheter from the right femoral vein. No arterial access was used for procedure monitoring. In the early period, sheaths were removed after activated clotting time (ACT) level dropped below 250 s and the puncture sites were compressed for 6 hours. Since 2014, the sheaths were removed at the end of the procedure and venous hemostasis was achieved by “Z”-stitch. Bed rest was implemented in all patients till the next morning.
Intracardiac echocardiography was used throughout the whole procedure as an institutional standard. It was used particularly for guiding the transseptal puncture, determining accurate antral ablation points, tagging the course of the esophagus, titration of radiofrequency energy, and early detection of complications.
CA was performed with a 3.5 mm irrigated-tip catheter (Navistar® Thermocool or Thermocool Smarttouch® or Celsius® Thermocool; Biosense Webster, Diamond Bar, CA, USA). The catheter was navigated with the use of 3D electroanatomic system (CARTO®; Biosense Webster or Ensite NavX®; St Jude Medical, St Paul, MN, USA). Linear point-by-point lesions were placed around the ostia of pulmonary veins with the endpoint of electrical isolation. In patients with the advanced atrial disease (38%), empirical linear lesions and/or superior vena cava isolation and/or biatrial electrogram-guided ablation was performed. Cavotricuspid isthmus was ablated in 22% of procedures. Mappable ATs were always targeted. Non-inducibility of arrhythmia was the desired endpoint of repeated procedures.
Radiofrequency energy was delivered by an EP Shuttle(Stockert, Freiburg, Germany) or Smartablate(Biosense Webster) or Ampere generator (St Jude Medical). Constant irrigation flow of 15 ml/min (30 ml/min inside the coronary sinus) through a Cool Flow® pump (Biosense Webster) or Cool Point (St Jude Medical), respectively, was employed. The power mode was used with a preset power up to 25–35 W and down-regulation when the tip temperature of 43°C was achieved. Power output was mostly reduced to 20–25 W at the left atrium posterior wall and inside the coronary sinus. The energy was delivered either in point-by-point fashion (20–30 s at one spot) or by dragging. Rarely (< 5%) other ablation catheters were used; of them predominantly PVAC or Arctic Front®(Medtronic, Minneapolis, MN, USA).
Perioperative anticoagulation
In the early period from 2006 to 2012, warfarin was temporarily discontinued to achieve International Normalized Ratio (INR) < 2 and bridged with weight-adjusted low molecular weight heparin (LMWH). At the beginning of every procedure, unfractionated heparin was administered with a loading dose of 5000 IU before the first transseptal puncture. The ACT was measured every 15–30 minutes, with the target value of 300–350 s in all patients during the procedure. After the procedure, the infusion of unfractionated heparin was administered until the morning of the first post-ablation day. Subsequently, the warfarin was restarted and LMWH administered until reaching the therapeutic level of INR.
From 2013, all procedures were performed on uninterrupted warfarin with target INR between 2 and 3 or minimally (single dose) interrupted direct oral anticoagulants. After the procedure, either direct oral anticoagulants or warfarin were re-started after the drop of the ACT under 170 s.
Anticoagulation treatment continued for at least 3 months after the procedure in low-risk patients or lifelong in high-risk patients according to the guidelines.5
Assessment of complications
The specific institutional tracking system was used to identify all complications during and after the procedure until the 3-month clinical outpatient visit. This included a purpose-established central registry for complications of invasive procedures, which was described elsewhere.6 All recorded complications were individually reviewed by staff physicians of the arrhythmia service in morbidity and mortality meetings and classified as vascular or other, and major or minor.
Definition of complications
Major complications were defined as those that result in permanent injury or death, require intervention for treatment, prolong (> 48 h) hospitalization or require new hospitalization.6 MVCs were major complications limited to vascular access. Hemoglobin drop by > 30 g/l was classified as MVC even if it was not treated with blood transfusion.
Statistical analysis
Continuous variables were expressed as means ± standard deviation and compared by t-test or Mann-Whitney U test, as appropriate. Categorical variables are expressed as percentages and compared by the chi-square test, Fisher exact test or logistic regression if appropriate. Association of baseline clinical and procedural factors with MVCs was investigated by linear regression analysis. Statistically significant factors on univariate analysis (P < 0.1) were entered into a multivariate regression model. Two-tailed α < 0.05 was considered statistically significant, except for the test of equality of covariance matrices where P < 0.005 was considered significant. For post hoc comparison of subgroups, Tukey HSD or Bonferroni test was used where appropriate. The majority of analyses were performed separately for males and females. All analyses were performed using TIBCO Statistica version 13.3 (Palo Alto, CA, USA) or IBM SPSS for MAC version 23 (IBM, New York, USA).