Atrial fibrillation (AF) is the most common arrhythmia present in 1% of population under 60 years of age and reaching up to 15% at 80 years. AF is associated with reduced quality of life, increased morbidity, mortality and health economic costs.

Presentation of AF differs substantially among patients ranging from self-limiting short episodes (paroxysmal AF), longstanding episodes (persistent AF) where direct current (DC) cardioversion is needed, to chronic atrial fibrillation. Treatment of AF is individually tailored in accordance to symptoms, type of AF and thromboembolic risk. The standard treatment of symptomatic persistent AF is DC-cardioversion preceded by anticoagulant treatment with Warfarin. According to guidelines DC-cardioversion can be performed when anticoagulation treatment has been in therapeutic range for at least 4 weeks. However introduction of Pradaxa (Dabigatran) has enabled an earlier DC cardioversion, reducing time to cardioversion to a 3 week period. During anticoagulation treatment persistence of AF contributes to left atrial remodeling and increases in inflammatory and neurohumoral biomarkers. The prolonged duration of AF and the remodeling of the left atrium increase the risk of AF recurrence after DC-cardioversion.

Early cardioversion of patients with persistent AF is possible if preceded by transesophageal echocardiography (TEE). The TEE guided DC- cardioversion, as demonstrated in the ACUTE study, is a safe and efficient alternative to conventional treatment. This treatment regime is not routinely used in clinical practice.

The aim of this study is to compare early DC-cardioversion (within 72 hours) to conventional treatment (Pradaxa prior to DC-cardioversion). 140 patients with persistent AF will be randomized to early cardioversion preceded by TEE in accordance with guidelines or conventional treatment with Pradaxa for 4 weeks prior to DC-cardioversion.

The investigators will determine the outcome in the two groups regarding:

• Left atrial function and size assessed by left atrial strain, left atrial ejection fraction and left atrial volume.
• Inflammatory and neurohumoral biomarkers including ANP, BNP,IL6 and CRP.
• Time to recurrence of AF (AF documented by ECG or Holter monitoring)

Comprehensive transthoracic echocardiography, 12 lead ECG, biomarkers and Holter monitoring will be performed at the time of randomization, 4 weeks, 3 month and 6 month post DC-cardioversion. Furthermore all patients will be followed for symptomatic AF recurrence for a period of one year. AF recurrence will be documented by 12 lead ECG.

Background:

Atrial fibrillation (AF) is the most common arrhythmia, present in about 1 % of the population under 60 years of age and significantly higher among the older population, reaching up to 15 % in patients at 80 years. Worldwide, the proportion of people aged over 60 years is growing faster than any other age group. This will inevitably result in an increased prevalence of AF in the overall population. This is supported by the fact that the incidence of AF has increased over the past two decades. Furthermore a number of diseases which are associated with a higher prevalence of AF, such as ischemic heart disease (IHD), valvular heart disease, diabetes and hypertension are also expected to increase in prevalence due to the aging population. This again will contribute to an increased incidence of AF. Atrial fibrillation is associated with significant increase in mortality, morbidity and a reduction in quality of life.

The mechanisms behind AF including proliferation of fibroblasts, enhanced connective tissue deposition and fibrosis, all play an important part in structural remodeling of the left atrium. The structural remodeling leads to electrical deterioration which is fundamental for initiation of AF. The persistence of AF induces further electrophysiological remodeling, shortening of atrial refractoriness and in the end persistence of AF. The vicious cycle can, be reversed by restoration of sinus rhythm.

The treatments available today range from DC cardioversion, pharmacological rhythm and rate control, radiofrequency ablation to anticoagulation treatment.

If rhythm control is desired, cardioversion to sinus rhythm is achievable by DC-cardioversion. If AF is present less than 48 hours DC-cardioversion can be performed without further precautions. If AF is present more than 48 hours, DC cardioversion should be preceded by anticoagulation treatment with Warfarin for a minimum of 4 weeks or alternatively Pradaxa for 3 weeks prior to DC-cardioversion. Early cardioversion of patients with persistent AF > 48 hours is possible if preceded by transesophageal echocardiography (TEE) and exclusion of thrombus as demonstrated in the ACUTE trial.

Discrepancy of the results and the limitations of the earlier studies lead to inconclusive results. Tendencies favouring early DC-cardioversion and an increase in haemorrhages in the conventional treatment group underline the need for a randomized study with a primary aim to asses' differences in effect of conventional vs. early DC-cardioversion treatment.

To the investigators knowledge no randomised study has assessed the effect of early versus late DC-cardioversion in patients with persistent AF.

Hypothesis:

Early DC-cardioversion of persistent AF will improve left atrial function and remodeling, decrease inflammatory /neurohumoral markers and increase time to recurrence of AF compared to conventional treatment.

Aims:

1. To determine, if there is difference in function and size of the left atrium, prior to and post cardioversion comparing early cardioversion to conventional treatment group. The data for comparison will be acquired echocardiographically.
2. To determine, if there is difference in levels of inflammatory (IL-6 & CRP) and neurohumoral markers (ANP & BNP) prior to and post cardioversion, when comparing early cardioversion to conventional treatment group.
3. To determine, if there is difference in time to recurrence of, ECG or Holter verified AF, when comparing early cardioversion to conventional treatment.

Descriptive aims:

1. To determine whether there is any difference in correlation between left atrial size and function and neurohumoral or inflammatory biomarker levels pre/post cardioversion, when comparing early cardioversion to conventional treatment group.
2. To determine, whether there is any difference in strain when analyzing TEE versus transthoracic echocardiography (TTE) images of the left atrium, in patient's receiving early cardioversion.

Methods :

Inclusion criteria

Patients admitted to department of cardiology OUH Svendborg Hospital or referred to outpatient clinic with symptomatic persistent AF, duration more than 48 hours and indication for DC cardioversion. Atrial fibrillation must be verified by a 12 lead ECG. All patients must be over 18 years of age, and must provide written informed consent prior to inclusion.

Exclusion criteria

Reversible causes for AF (thyrotoxicosis, infection, pulmonary embolism), acute coronary syndrome and absolute contraindications of TEE (oesophageal spasm, stricture, perforation and diverticula). Patients with diminished mental capability will not be included.

Study design

The patients will undergo unblinded randomization in two groups with PROBE design which includes blinded analysis of echocardiography, Holter monitoring and biomarker results.

Group one will receive conventional treatment with Pradaxa for a period of minimum 3 weeks prior to DC cardioversion. Group two will undergo TEE followed by early DC cardioversion within 72 hours. Post cardioversion treatment with Pradaxa will be applied according to guidelines for a minimum of 4 weeks.

Comprehensive transthoracic echocardiography, 12 lead ECG, biomarkers and Holter monitoring will be performed at the time of randomization, 4 weeks, 3 month and 6 month post DC-cardioversion. Furthermore all patients will be followed for symptomatic AF recurrence for a period of one year, AF recurrence will be documented by 12 lead ECG.

Procedures :

Transesophageal echocardiography Patients will be examined in the fasting state, after the application of topical pharyngeal anaesthesia. Left atrium and left atrial appendage will be viewed in order to exclude thrombus and dynamic spontaneous echo contrast. Tissue Doppler and 2D images of the left atrium from 0,30,60 and 90 degrees will be acquired for the offline strain analysis.

Transthoracic echocardiography Two dimensional images and tissue doppler images (TDI) of the left atrium and the left ventricle will be acquired for offline strain analysis. Standard evaluation of systolic function (EF, WMSI); diastolic function (E/a, DT, e', s', atrial volume) and evaluation of mitral and aortic valves will be performed. Blinded offline strain analysis will be performed with commercially available software. "Time to peak standard deviation"(TPSD)a reliable measurement of overall left atrial dyssynchrony, peak strain and global atrial strain will all be computed during the offline analysis.

Blood specimen Will be drawn prior to cardioversion and 4 weeks post cardioversion. The blood samples will be stored immediately after the collection in a - 80°C refrigerator. Neurohumoral and inflammatory markers (IL 6, ANP, BNP, CRP, ect.) will be analyzed en bloc.

Holter monitoring /12 lead ECG 48 hour Holter monitoring and a 12 lead ECG will take place 4 weeks, 3 month and 6 month post cardioversion. A recurrence of AF will be defined as AF for a period of minimum 30 seconds on Holter or AF verified by 12 lead ECG.

Sample size Based on previous studies of changes in peak strain following DC cardioversion the investigators expect a mean change in peak strain of 5.8 in the control group. The investigators expect a minimum relevant difference (MIREDIF) between the two groups of ½ standard deviation ( 4.05 %). With an α of 0.05 and a β of 0.80 the investigators must include a minimum of 63 patients in each treatment group. With an expected dropout of 10% the total sample size is 140 patients included. A thorough review of medical records for a period of 01/01/2010 - 01/10/2010 show that DC-cardioversions has been performed 121 times at OUH, Svendborg Hospital during a period of 10 month. Based on the number of previous DC-cardioversions the investigators expect an inclusion period of approximately 18 month.

Data Analysis Outcome and explanatory variables will be tested for normality and transformed when appropriate. Univariate analyses of changes in atrial strain and neurohumeral response between the two groups will be performed using unpaired t-test and simple regression. Multivariate regression analysis will be used to analyze significant covariates. Analysis of repeated measures will be performed in a Mixed Model approach. Time to recurrence will be analysed using log-rank test and Cox Proportional Hazards Model.

Ethical Considerations This clinical trial will be performed in accordance with ethical principles based on the Declaration of Helsinki and will be approved by the ethics comity. Prior to inclusion all patients will receive a thorough explanation of this trials nature; participation will be based on signing of an informed consent. The patients may, at any given time withdraw from the trial. The trial itself will be approved by ethical and scientific counsels. Furthermore both patient groups will receive an already established and approved treatment, and patients will therefore not be submitted to experimental treatments and/or risks that such treatments may present.

Inclusion Criteria:

• Patients admitted to department of cardiology OUH Svendborg Hospital or referred to outpatient clinic with symptomatic persistent AF, duration more than 48 hours and indication for DC cardioversion. Atrial fibrillation must be verified by a 12 lead ECG. All patients must be over 18 years of age, and must provide written informed consent prior to inclusion.

Exclusion Criteria:

• Reversible causes for AF (thyrotoxicosis, infection, pulmonary embolism), acute coronary syndrome and absolute contraindications of TEE (oesophageal spasm, stricture, perforation and diverticula). Patients with diminished mental capability will not be included.