The main aim of this project is to assess the effect of Sleep apnea syndrome (SAS) screening (and treatment if SAS is moderate to severe, defined by an apnea hypopnea index >15 / h) on recurrence of atrial fibrillation (AF) over a twenty-four month follow-up period, in patients on optimal medical treatment after AF ablation.

Sleep apnea syndrome (SAS) is a common but often undiagnosed disorder associated with substantial cardiovascular morbidity and mortality. SAS prevalence in atrial fibrillation (AF) is 20% to 75%. SAS is associated with both myocardial fibrosis and electrical remodeling which both favor recurrence of AF despite medical treatment or atrial ablation. Proportion of patients with SAS are higher in patients with high-frequency paroxysmal AF and persistent AF than those with low frequency paroxysmal AF. Furthermore, untreated SAS doubles the risk of recurrence of AF after electrical cardioversion. Noteworthy, up to now, given the large number of potential candidates, not all patients with AF are screened for SAS despite recent international guidelines.

Observational studies suggest that treatment of SAS can prevent recurrence of AF episodes but no randomised trial demonstrate this with an adequate level of evidence. In addition, SAS treatment with continuous positive airway pressure (CPAP) has proven its efficacy on other pathologies on cardiovascular outcomes as stroke and to a lesser extend hypertension. In several countries including France, CPAP is reimbursed by national health services. However there are few data on the cost effectiveness of SAS treatment from the standpoint of cardiovascular health costs and no study in AF. Limits of small randomized studies and observational cohorts are fairly illustrated by very recent results of a randomized trial, the SERVE-HF study, which raised the question of central SAS treatment on mortality in systolic heart failure patients. This multicentre randomized trial showed no difference in global mortality between groups, with even an increased OR for cardiovascular death in patients treated for central SAS. In this context, the need of a randomized trial in AF, with a data safety and monitoring board, is even more important to confirm results of observational studies.

Our hypothesis is that SAS screening (and SAS-treatment if screening is positive) in patients presenting with recurrent AF will decrease the risk of recurrence of AF, therefore decreasing associated health costs.

SAS is largely under-diagnosed in AF patients and screening is recommended despite lack of high level-evidences on efficacy of SAS treatment on AF recurrence, and still low resources in terms of SAS screening and absence of cost-effectiveness analysis.

The investigators propose a study design in which patients are randomized for SAS screening in two groups :

• the control group will receive usual care and follow-up of AF;
• the intervention group will benefit of screening for SAS and treatment if relevant (moderate to severe SAS defined by an apnea hypopnea index (AHI)> 15/h), in addition to usual treatment of AF.

The cost-effectiveness of a systematic screening for SAS in AF has never been studied despite the personal and economic burden of SAS treatment. The investigators will analyze the consequences on direct medical costs of systematic SAS screening and treatment in the patients with AF and whether SAS therapy would decrease healthcare resource utilization and costs, via decreasing AF recurrence.

The main aim of this project is to assess the effect of SAS screening (and treatment if SAS is moderate to severe, defined by an apnea hypopnea index >15 / h) on recurrence of AF over a twenty-four month follow-up period, in patients on optimal medical treatment after AF ablation.

A 2 group parallel, randomized (1:1 ratio), Prospective Open Blinded Endpoint (PROBE), multicentric controlled trial of superiority.

The control group will receive usual care and follow-up of AF. The intervention group will benefit of screening of SAS and treatment if relevant, in addition to usual treatment of AF.

To minimize cross over from the control to the intervention group, patients of the control group will performed a polygraph at the end of the 2 year follow up. Both groups will be stratified by center.

SAS diagnosis will be performed by polygraphy scored using international criteria (American Academy of Sleep Medicine 2012). If AHI > 15/h, treatment will be started, with continuous positive airway pressure (CPAP) or mandibular repositioning device (MRD) based on type and severity of SAS.

Both groups will be followed over 24 months by the cardiologist, and by the somnologist / pulmonologist for the patients treated for SAS.

Inclusion criteria :

• Patients experiencing at least a second episode of symptomatic and documented atrial fibrillation (paroxysmal or persistent) within the preceding 12 months, unrelated to an acute cause, considered to be appropriate candidates for rhythm-control therapy (catheter ablation therapy)
• Age >18 y and < 80y
• Expected hospitalization for atrial fibrillation ablation

Exclusion criteria :

• Already known SAS,
• Severe sleepiness (Epworth sleepiness score > 14)
• Driving professional
• Severe Chronic obstructive pulmonary disease (COPD)
• Any other illness susceptible to shorten life expectancy within the study duration.
• Previous ablation for atrial fibrillation,
• Secondary atrial fibrillation (due to cardiac surgery, infection, or hyperthyroidism).
• Severe heart failure (NYHA functional class III to IV),
• Expected surgery for structural heart disease
• Unable to understand and to comply to clinical research procedures,
• Not covered by the national health insurance system
• No consenting for a randomised controlled trial
• Previous exam missing
• Pregnancy and breast feeding
• Guardianship or trusteeship patient