Catheter Ablation Compared With Pharmacological Therapy for Atrial Fibrillation (CAPTAF Trial)

It is a randomized, open-label, multi-centre trial in which patients with symptomatic paroxysmal or persistent atrial fibrillation will be randomized to treatment with either catheter ablation for pulmonary vein isolation or optimized conventional pharmacological therapy. Patients will receive an implantable cardiac monitor, Reveal XT, implanted subcutaneously, before a two month "Run-in" period from which the basic atrial fibrillation burden will be defined, which will provide an objective basis to assess the AF burden for 3 years.

After initiation of randomized therapy, patients will be followed at 3, 6, 9, 12, 18, 24, 30, 36 and 48 months. In case of intolerable symptomatic atrial fibrillation despite allocated treatment, subjects will be switched to the alternative regimen, on patient's request.

The ablation will consist of pulmonary vein isolation for both paroxysmal and persistent atrial fibrillation. A linear lesion at the left atrial roof is optional for patients with recurrence after a first procedure or primarily for patients with persistent atrial fibrillation only. A second ablation procedure may be undertaken at earliest 3 months after the first procedure, if symptoms persists or recurs. Epicardial off-pump ablation may be offered after failed attempts of transvenous catheter ablation at earliest 3 months after latest procedure. If the second transvenous ablation procedure fails, patients may be offered i) a third transvenous ablation procedure or ii) an epicardial off-pump procedure. In case the patient declines a third intervention despite recurrence, the patient should continue on the same anti-arrhythmic drugs. If new previously untested drugs are prescribed, the patient will be defined as a change of treatment (and failed ablation). Ablation may be performed using radiofrequency energy with irrigated tip or with cryo-energy using the Arctic Front™ Cardiac CryoAblation Catheter. A complete pulmonary vein isolation will be defined by documenting PV entrance block in each vein using a circular mapping catheter.

Anti-arrhythmic drugs remaining to be tested for rate and rhythm control may include previously tested tolerable antiarrhythmic drugs if dosages were inadequate. Optimized anti-arrhythmic drug therapy includes testing of all available anti-arrhythmic drugs at adequate dosages, including amiodarone 600 mg once daily for 7-10 days, and 100-200 mg once daily thereafter; sotalol: 80-160 mg twice daily; flecainide 100 to 150 mg twice daily or the entire dose as slow-release formula once daily; propafenone 300 mg twice daily; and disopyramide 250-375mg twice daily, and potentially new antiarrhythmic drugs commercially available during the study period.

A monitor will verify that data recorded on the procedure forms are correct through quarterly communication, review of catheterization reports and medical records. In accordance with applicable regulations a monitor (study nurse from Uppsala) will contact the site prior to the start of the study to review with the site staff protocol, study requirements and their responsibilities to satisfy study requirements, if needed. When reviewing data collection procedures the discussion will also include identification, agreement and documentation of data items for which the case record forms (CRF), patient files (paper or database), tracings of investigations, and questionnaires, will serve a source document. The investigator and the head of the medical institution agrees should allow the monitor direct access to all relevant documents whenever needed and in an event of an audit. To ensure compliance the monitor may conduct a quality assurance audit, which can occur at any time during or after completion of the study.

The Biometrics section at Uppsala Clinical Research Centre is responsible for the Data Management. All data will be recorded in the CRFs and entered via eCRF directly into a web based data capturing system except for electronically available data that will be loaded directly into the study database. Each investigational site will have authorized site personnel responsible for entering the data, as well as for changing and correcting the data according to instructions provided by Uppsala Clinical Research Centre (UCR) and the coordinating centre. All changes will be tracked by an audit trail. The site specific investigator will sign the eCRF electronically when the data have been reviewed and edited, and Source Data Verification has been performed.

The Database Closure will be performed in three steps, the first one when all data from the first 12 months are entered, the second one when all data from the 24 months period are entered and the third when the complete study period of 48 months are entered. All datasets used for the 12- and 24 months analyses will be locked separately. Procedures: When all data are entered into the database and all queries solved, the Database Closure procedures will start.

Complications are defined as acute if occurring within 1 week after initialized therapy, and late if occurring one week after initialized therapy. All "Serious adverse events" should be documented on the "Adverse events Form" AND "Serious adverse events Form". All " Adverse events" should be documented in the "Adverse events Form". All events which are part of performing a catheterization procedure will be reported. All serious unanticipated events whether or not they are considered to be related to the procedure but which have some association with the catheterization either by timing or physiology, will be reported. Other bleedings than major bleeding events should be reported as minor bleeds. All events related to pharmacological therapy will be reported.

Procedural complications are defined as those arising during or within 1 month of the procedure, including death, major bleeding, vascular injury, cerebrovascular accident, myocardial infarction, pericardial effusion with or without tamponade, venous or systemic embolism, phrenic nerve paralysis, and heart block. Events occurring later than 1 month after the procedure and defined as related to the procedure are pulmonary vein stenosis, tamponade, retroperitoneal haematoma, and atrio-esophageal fistula (see below).

Relation to Device Procedure or drug, Intensity, Outcome and Actions taken will be described according to international regulations.

Statistical methods. All continuous variables will be presented per treatment group using descriptive statistics by mean, SD, max and min values, in addition medians, 25th and 75th percentiles will be presented when suitable.

The analysis of mean change in the primary endpoint will be two-sided and performed as an unpaired t-test at 5% significance level. The difference between the treatment groups with corresponding two-sided 95% confidence interval based on a normal approximation will be presented. The continuous secondary endpoints will be analyzed in the same way as the primary endpoint, but the results will be interpreted descriptively. Kaplan-Meier estimates and log-rank test will be used to determine the occurrence of the secondary endpoints over time. All categorical secondary endpoints will be compared between treatment groups with frequency tables and Fisher's Exact Test including 95% confidence intervals where possible. Adverse Events will be summarized per treatment group by body system/system organ class, preferred terms, intensity, seriousness and relationship. Laboratory data for efficacy will be presented in mean change tables.

The null hypothesis (H0) is that the difference in the mean change from baseline to 12 months after baseline in General Health between the two treatments is zero;- mean change for the treatment group catheter ablation (μ 1) and - mean change for the treatment group optimized conventional pharmacological therapy (μ 2) Safety - All randomized patients will be included in the safety analysis. Only observed observations are used in the safety analysis.

Intention to treat (ITT) - All randomized patients. The main analysis will be performed on the ITT-population.

Per protocol (PP) - All randomized patients completing the study treatment period of 12 months without any major protocol violation (for example ineligibility, early withdrawals, poor compliance).

A subgroup analysis will determine whether sinus rhythm obtained by AF ablation is superior to sinus rhythm obtained by pharmacological therapy for rate and/or rhythm control. In addition a comparison will be done between patients with sinus rhythm and patients with atrial fibrillation irrespective of allocated therapy. Both analyses will be performed with regard to quality of life, morbidity (composite, see secondary endpoint 1), cardiovascular hospitalizations, exercise/ physical capacity, safety, and health economy at 12 and 24 months.

All continuous variables will be presented per treatment group using descriptive statistics by mean, standard deviation, max and min values, in addition medians, 25th and 75th percentiles will be presented when suitable. The analysis of mean change in the primary endpoint will be two-sided and performed as an unpaired t-test at 5% significance level. The difference between the treatment groups with corresponding two-sided 95% confidence interval based on a normal approximation will be presented.

The continuous secondary endpoints will be analyzed in the same way as the primary endpoint, but the results will be interpreted descriptively.

Kaplan-Meier estimates and log-rank test will be used to determine the occurrence of the secondary endpoints over time. All categorical secondary endpoints will be compared between treatment groups with frequency tables and Fisher's Exact Test including 95% confidence intervals where possible. Adverse Events will be summarized per treatment group by body system/system organ class, preferred terms, intensity, seriousness and relationship.

Sample size considerations: The primary endpoint is the change in General Health from baseline to 12 months after baseline. Based on previous studies it is assumed that the variable is normally distributed with a standard deviation for the change in General Health of 20 units and an expected difference between groups of at least 10.5 units. To detect a difference of 10.5 units in General Health (corresponding to an improvement of 15%, assuming a mean General Health of 70 units in the conventional pharmacological therapy group) with a power of 80% and a type I error of 5% (two-sided alternative), a sample size of approximately 58 subjects in each treatment group is required, i.e. a total number of 116 patients. Another 20 patients are added to ensure calculated number of patients for analysis at 12 months. The power to detect a difference of 10.5 for 136 (116+20) and 140 (the minimum expected total number) patients is 85% and 86%, respectively.

A 1:1 block randomization will be used within each centre, stratifying patients between paroxysmal and persistent AF.