Reverse Remodelling of the Atria After Treatment of Obstructive Sleep Apnoea with Continuous Positive Airway Pressure: Evidence from Electro-mechanical and Endocrine Markers

BACKGROUND:Recent studies have suggested a strong association between obstructive sleep apnoea (OSA) and atrial fibrillation (AF). However, the impact of continuous positive airway pressure (CPAP) therapy on reverse atrial remodelling in patients with OSA is poorly understood. We aimed to determine the impact of CPAP therapy on total atrial conduction time measured by PA tissue Doppler imaging (TDI) interval in patients without history of AF. METHODS AND RESULTS:Thirty-five patients with first diagnosed OSA (group 1) and 20 reference patients without OSA (group 2) were prospectively enrolled. Group 1 was divided into 25 patients with starting CPAP therapy (intervention group) and 10 patients without CPAP therapy (control group). PA-TDI interval, measured by tissue Doppler imaging, and B-type natriuretic-peptide (BNP) were determined at baseline and repeated after 30 days in group 1. Patients with OSA had a significantly longer PA-TDI interval (131.4±16.0 ms vs. 120.1±6.4 ms, p<0.001) compared to reference patients. PA-TDI interval shortened significantly after 30 days of CPAP therapy (131±17.1 ms vs. 124.6±15.7 ms, p<0.001), whereas no significant change was found in the control group (132.6±13.6 ms vs. 133.2±13.1 ms, p=0.31) and BNP-levels decreased significantly (39.1±50.7 pg/ml vs. 28.2±42.4 pg/ml, p<0.01), while BNP-levels remained unaffected after 30 days in control group (31.5±20 pg/ml vs. 34.4±20.5 pg/ml, p=0.41). CONCLUSIONS:Change in PA-TDI interval, as an indirect surrogate marker of atrial remodelling, suggests reverse atrial remodelling in patients with OSA and treated by CPAP. This effect may ensue from decreased atrial pressure or volume overload, as indicated by significantly reduced BNP levels. These observations suggest that the substrate pre-disposing to AF may be reversible and moreover can be measured by PA-TDI interval and BNP.