The aim of the present prospective study is to further develop and validate a composite risk score predicting both ischemic and bleeding risk, based on epidemiologic, clinical, biological, and/or morphologic complementary data. First, the investigators will assess the predictive performance of current clinical risk scores. Second, the investigators will assess the potential predictive value of additional markers. Third, the investigators will aim to develop a new risk score.

Patients with Atrial Fibrillation (AF) have an excessive risk of arterial thromboembolic disease and heart failure, resulting in higher rates of cardiovascular and all-cause mortality in this population.

Prevalence and incidence of Non-Valvular AF (NVAF) is rapidly increasing across developed countries. cardiovascular (CV) and all-cause mortality rates related to AF have increased between 1990 and 2010, irrespective of sex.

Determinants of AF (heart disease and/or cardiovascular risk factors), are themselves associated with an excess risk of cardiovascular events.

Primary prevention of NVAF relies on the control of modifiable risk factors, such as overweight, arterial hypertension, diabetes, and tobacco consumption which are associated with the risk of AF risk and its complications.

Preventive treatment of cardiovascular complications is based on anti-thrombotic treatment, which reduces the global ischemic risk but at the expense of an increase of bleeding risk.

Treatment decisions are guided by the evaluation of ischemic and hemorrhagic risks, as determined by clinical scores (like CHADS2; CHA2DS2-VASc; HAS-BLED) that have been validated on varying populations through retrospective cohort studies. However, since their publication, these clinical scores have been criticized regarding their predictive value, the lack of precision in the definition of their component and the threshold defined for treatment decisions. Furthermore, several studies have demonstrated the interest of biomarkers in addition to clinical parameters to predict the risk of cardiovascular complications in AF. For example, preliminary data suggest that the dosage of Brain-type natriuretic peptide (BNP), N-terminal (NT) Pro-BNP, troponin, C-reactive protein (CRP), and all indicators whose levels increase independently of their usual causes may contribute to a better prediction of cardiovascular complications including all-cause mortality in NVAF. Of note, the rate of major bleeding events was very low (1.5 per cent among 3,978 patients from the Euro Heart Survey on AF with complete follow-up) in the HAS-BLED validation cohort.

We hypothesize that epidemiologic, clinical, biologic, and morphologic complimentary data could improve the stratification of cardiovascular risk in NVAF. Biologic and echocardiographic approaches, in particular, could improve the performance of routinely used clinical scores. Consequently, identification of a state of inflammation, hypercoagulability, or increased circulating concentrations of certain biomarkers could explain the increased risk of cardiovascular events in AF. Furthermore, the morphologic repercussions of NVAF, including left atrium dilation and left ventricular systolic dysfunction, as well as the presence of left atrial appendage stasis indicators, could also help refine the risk stratification.

In a previous retrospective survey, the investigators have shown that biomarkers and echomarkers could better stratify AF patients at increased cardiovascular risk. However, due to its retrospective design and the lack of completeness, the investigators could not validate a new risk score including all these markers and taking into account both thromboembolic and hemorrhagic risk. Indeed, these usual biologic and morphologic parameters will be obtained at admission in all patients in a prospective design.

There are three study sites.

Professor Ariel COHEN is the coordinating investigator of the study. A Clinical Research Associate working within the Cardiology Service of the Saint-Antoine Hospital will monitor the study.

The study duration is 5 years:

• Inclusion period: 3 years
• End of study: 2 years after the end of inclusion period
• Follow-up: the follow-up period will run for 2 years.

Inclusion: The inclusion visit will occur during the patient's hospitalization for NVAF in the Cardiology Department. At the inclusion visit, information will be collected regarding the patient's history of clinical events, current risk factors and treatments.

Follow-up: After hospital discharge, each patient will be seen every six month throughout the two year follow-up period.

During the follow-up period, information on clinical events, changes in risk factors and treatments occurring since the last follow-up contact will be collected for each patient, a blood sample will be collected, and an echocardiography will be conducted.

Classification of preceding events and follow-up information will be supported by both clinical exams and medical file evaluation. Interviews with general practitioners and the patients themselves will also be used in order to collect all necessary information. In the event of hospitalization, hospitalization reports will be collected. In cases of events without hospitalization, general practitioner correspondence and prescription records will be collected.

Patients with NVAF will receive anti-thrombotic treatment, in accordance with routine care and European society of cardiology (ESC) guidelines.

The choice of anti-thrombotic treatment will be entirely at the discretion of the attending physician during hospitalization.

Information describing all treatments received by each patient will be collected throughout the study, from the inclusion visit to the final follow-up contact.

All echocardiographic procedures will be performed in the echocardiographic laboratory of each Cardiology Department. Cardiologists specialized in echocardiographic examination will perform all echocardiographic procedures.

Each participant will undergo an echocardiographic examination at their inclusion and at each follow-up visit occurring every 6 months. The objective of the echocardiographic examinations is the identification of cardiac conditions and diseases that may be associated with AF.

Primary Outcome variables The primary outcome variable is a composite cardiovascular endpoint including stroke, transitory ischemic attack, thromboembolism, acute heart failure, acute coronary syndrome, major bleeding and cardiovascular and all-cause mortality.

(Major bleeding will be defined according to international society on thrombosis and haemostasis (ISTH) criteria)

Secondary Outcome variables

The incidence of each of the following events will be determined:

• death (all-cause and cardiovascular)
• acute heart failure
• stroke
• acute coronary syndrome
• major bleeding episodes

Enrolled participants will be included in the analysis. Participants who withdrew their consent will be excluded from any analysis. Participants who had no event or those who were lost to follow-up (impossibility of contact for more than 24 months) will be included up to the date of their last contact.

Event free survival is defined as the time period from the day of enrollment in the study to the day of the event or death.

Overall survival will be calculated for each patient starting at their enrollment until their death or the end of follow-up. One analysis per event type will be conducted.

One vital status research will be conducted for each participant lost to follow-up; where appropriate a cause of death research will be undertaken. To deal with the possibility for a same patient to present multiple events, we will consider competing risk models.

Inclusion Criteria:

• Adults above 18 years of age
• Presenting a NVAF which is paroxysmal, persistent or permanent and documented by an electrocardiogram (ECG)
• Willing and able to give written informed consent

Exclusion Criteria:

• Patients presenting a clinically and echocardiographically significant valve disease defined as :

  • mitral stenosis with a valve area < 2cm2
  • mitral insufficiency (grade 3 or 4)
  • aortic insufficiency (grade 3 or 4)
  • tricuspid insufficiency (grade 3 or 4)
  • aortic stenosis with a valve area < 1.5 cm2
  • mechanical valve prosthèses

• Secondary AF due to an acute cause (thyrotoxicosis, acute pericarditis, infection, systemic disease)

• Presence of a contraindication to long term anticoagulant treatment

• Absence of social security coverage

• Severe psychiatric history

• Impossibility of follow-up

The study population will be composed of patients hospitalized in the Cardiology/Rythmology Department of the recruitment sites for an episode of non-valvular atrial fibrillation during the inclusion period.

each participant can have access to its data by asking the investigator of the study.