Electrical Cardioversion of Recent Onset Atrial Fibrillation - Silent Thromboembolic Events, Reverse Atrial Remodeling

Working plan This study will give information about the incidence of silent cerebral thrombo-embolic events in patients with recent onset AF, a population which is expected to consist mostly of paroxysmal AF patients.

Patients with atrial fibrillation (AF) duration less than 48 hours and fulfilling inclusion and not exclusion criteria will be included in the study by a cardiologist on the day of cardioversion, and study nurse will be notified to plan for the investigations. The patient will undergo clinical examination, clinical history will be taken and blood sampling. Once 12 lead ECG, echocardiography, questionnaires, telemetry and MR have been done the patient will be electrically cardioverted After the cardioversion the patient will be subject to echocardiography of the heart again, blood-sampling, 12 lead ECG, and MR brain. The patient is discharged but will be studied by echocardiography of the heart and MR brain again on day 7 - 10. Thereafter the patient will be followed for 30 days - see flowchart.

The day of cardioversion is defined as Day 0.

3.1. ASSESSMENT OF REMODELING/REVERSE REMODELING 3.1.1. Electrical remodeling Standard 12-Lead electrocardiography (ECG) at every health care visit. (Day -1, Day 0 before and immediately after DCC, Day 0 at discharge, Day 7-10 and 30). Recorded with automatic analysis of intervals, Paper speed 50 mm/sec.

O Data to be collected: Rhythm, heart rate, P wave duration and amplitude.

3.1.2. Functional remodeling (Echocardiographic analysis) - App. 1. Left and right atrial dimensions and volumes 53. Global left atrial function, left atrial ejection fraction: (assessed by 2D and speckle tracking )53, 54 Left ventricular ejection fraction and left ventricular diastolic function (transmitral velocities, E/E' index)53-56 O Echocardiographic data will be collected and stored in the routinely used database of the department of Clinical Physiology and as raw images in a separate database dedicated for the study (in the care of the investigators) for offline review and measurements. The Core lab in Uppsala will do all analysis.

O Time of collection: prior to cardioversion, within 24 h after cardioversion and at day 7 - 10 after cardioversion.

3.1.3.Neurohormonal - inflammatory indices, cardiac biomarkers, vasoactive peptides.

High-sensitivity cardiac troponin T (hsTnT) and N-terminal pro-brain natriuretic peptide (Nt-proBNP) will be measured as sensitive and specific markers of cardiac injury ⁄ strain ⁄ filling pressures and B-type natriuretic peptide levels correlate with AF burden in patients with lone AF and is a strong predictor of recurrent arrhythmia after ablation. (A stable fragment of vasoactive peptide, C-terminal -proendothelin-1 will be measured because of its relation recurrence of AF, which may be a risk factor for embolism and because of its possible role in the pathogenesis and recurrence of AF, and the association with atrial dilatation, fibrosis, and hypertrophy, which may contribute to AF persistence and embolism) C-Reactive protein (CRP) Interleukin-662-65 Routine blood tests: (Hb, Platelets, White blood cell count, Na, K, Creatinine, international normalized ratio (INR), thyroid-stimulating hormone (TSH), free-T4): only once before cardioversion.

O Blood sampling test: Day 0 before (sample 1) and 4 + 1 hours after cardioversion at time for MRI no 2 before discharge (sample 2) and day 7-10 (sample 3). The core lab in Uppsala will do all analysis.

3.2. RHYTHM MONITORING O Continuous ECG monitoring O During and after cardioversion to be printed from cardioverter or telemetry or 7 - 10 day Holter.

O 7-day Holter monitoring. To be initiated on Day 0 prior to cardioversion for 7 days. The Core lab in Uppsala will do all analysis.

12 lead ECG recording As above and at symptoms indicating recurrence of AF.

Data to be collected:

Time to first P-wave to assess sinus node recovery after DC and time for immediate recurrence of AF.

AF burden (defined as total time in AF during the first 7 days) Time to first AF recurrence (sustained AF; duration >30 seconds during first 7 days, then at first symptom of AF confirmed on ECG) Number of sustained and non-sustained AF episodes during the first 7 days Mean, median - range of duration of AF episodes during the first 7 Days

3.3. ASSESSMENT OF THROMBOEMBOLIC BIOMARKERS, MENTAL TESTS AND CEREBRAL ISCHEMIC EVENTS

3.3.1. Biomarkers Coagulation activity by analysing plasma markers for thrombin activity (P-selectin, d-dimer, and prothrombin fragment 1+2, von Willebrand factor antigen (vWF-Ag)) and factor VIII:C.

Active fibrinolysis (plasmin-alpha 2-plasmin inhibitor complex: PIC) P-fibrinogen och P-Fibrin Platelet activity (platelet factor 4: PF4). Brain damage markers: S100 Blood sampling test as above.

3.3.2. Neurological / Cognitive assessment National Institute of Health Stroke Scale (NIHSS): Standardized assessment of neurological deficit66.

To be performed: Day 0 prior to cardioversion, Day 7-10, Day 30 and when clinically evident cerebrovascular event.

Mini-Mental-Test (Mini Mental State Examination, MMSE) and Trail Making Test A och B (TMT A och B) To be performed: Day 0 prior to cardioversion, Day 7-10, Day 30 and if clinically evident cerebrovascular event.

3.3.3. Brain magnetic resonance imaging. MRI is performed for the assessment of silent thromboembolism, which is defined as 2-3 mm or larger lesion with restricted diffusion on MRI with diffusion weighted sequence.51,71-75 Silent or clinically evident cerebrovascular events occurring during cardioversion will be detected and eventually not counted as late coming emboli in relation to reverse remodeling.

To be performed: Prior to and within 24 hours after DCC, on Day 0, and on Day 7 - 10 after DCC.

Data to be collected: presence, size, number, and vascular distribution of any focal abnormality consistent with embolic lesion.

Technique for MRI without contrast injection:

Sag T2 or T1 to position and obtain transversal slices with high reproducibility.

"DWI" transversal, 5 mm slices, b=0 och b=1000, calculation of "ADC" maps, usually automatic by scanner T2 tse/fse transversal, 5 mm slices T2 FLAIR transversal, 5 mm slices

1. Statistical analysis, sample size, data handling This pilot study will give information about the incidence of silent cerebral thromboembolic events and associated clinical variables that may affect the outcome, in patients with recent onset AF who are expected to mainly suffer from paroxysmal AF.

Based on previous reports of patients undergoing MRI before and after AF ablations, the % of patients with chronic cerebral lesions detected on MRI at baseline varies widely from 4 to 79 %, with 10 % being the most representative figure for patients with paroxysmal AF. Previous observations have reported varying frequencies of new silent cerebral lesions on MRI after AF ablation (4-38 %), coronary angiography (10 %) and retrograde aortic catheterisation of patients with aortic valve stenosis (22 %). New asymptomatic cerebral ischemic lesions in patients undergoing AF ablation have been reported after the use of irrigated radiofrequency (RF) catheters (7-11 %), the most common routine ablation procedure, and even a higher frequency of lesions (37 %) using a specialized circular RF ablation catheter. Based on these observations the investigators argued that a 20 % increase in incidence of new asymptomatic cerebral ischemic lesions in patients undergoing DC cardioversion would be clinically significant and warrant alternative routines for cardioversions.

Determination of sample size Considering a 20% increase (i.e. from 10% at baseline to 30%) in incidence of silent cerebral lesions following electrical cardioversion, the required number of patients to reject the null-hypothesis with 90% power and at the 0.05 significance level (two-sided) is 35. To allow for drop-outs, 40 patients will be included in the study. The sample size calculations were performed in the software "nQuery Advisor" version 7.0.

It is reasonable to perform a pilot study of 40 patients, in order to get an indication of the number of silent cerebral events. A total of 70 patients will be screened to ensure that 40 patients will remain in sinus rhythm at least 7 days after cardioversion.

Categorical variables will be reported as counts and percentages, whereas continuous variables as means and standard deviations (SD) or medians and interquartile ranges, as applicable.

The relationship between clinical variables (qualifying history duration of AF, AF duration, CHADS2VASC score, biomarkers, conversion to sinus rhythm, left atrial function and left atrial volume), confounding variables and incidence of new silent cerebral ischemic lesions at Day 1 and 7-10 post MR will be tested in a multiple logistic regression model, using a stepwise selection procedure (p<0.05). The model will be validated using bootstrapping.

The number of new silent cerebral ischemic lesions will be modelled as poisson or negative binomial regression models, as applicable.

Median concentrations of the biomarkers will be compared with baseline levels by nonparametric Wilcoxon rank-sum test or Kruskal- Wallis test when appropriate.

The association between biomarker concentrations and new embolic lesion will be evaluated over a follow up period 7 -10 days. The association between baseline concentration of each biomarker and new embolic lesion will be assessed by univariable Cox proportional hazards models. Biomarkers will be modelled as continuous variables (expressed as 1 SD increment) as linearity of the hazard will be tested by appropriate transformation using restricted cubic splines to account for possible nonlinear relationships. Cox multivariable models will be performed to evaluate the independent prognostic value of each biomarker separately; adjusting for baseline covariates emerged as statistically significant from a stepwise selection procedure (p < 0.05). The investigators will also investigate whether the addition of different combinations of the biomarkers improve the discrimination of the model. The independent prognostic value of each biomarker on new embolic lesion will be assessed by univariable and multivariable Cox models.

Each biomarker measured at baseline, Day 1 before and immediately after DCC, at discharge, after 7 - 10 days, and 1 months will be analysed graphically over time at the patient level as well as by summary statistics.

All probability values are two-tailed, and 95% confidence intervals (CIs) will be calculated. Because of the exploratory nature of this study, adjustments for the multiplicity of statistical analyses will not be made.

A 2-sided P-value <0.05 will be considered statistically significant.

1. Clinical Significance It is of paramount importance to assess if electrical cardioversion performed within 48 hours after AF onset, without preceding oral anticoagulation, results in silent strokes, as it is the present clinical routine and no scientific data is available.

A 20 % increase in incidence of new asymptomatic cerebral ischemic lesions in patients undergoing DC cardioversion is in this study defined as clinically significant. If other cardioversion strategies can be identified that result in 10 % lower incidence of events, that figure may then be regarded as clinically significant.