Effect of oxygen versus adaptive pressure support servo-ventilation in patients with central sleep apnoea-Cheyne Stokes respiration and congestive heart failure

Abstract

Background and aims: Central sleep apnoea with Cheyne-Stokes respiration (CSA-CSR) is a common, serious consequence of congestive heart failure. Optimal treatment is yet to be established. We compared two common treatments for CSA-CSR.

Methods: Subjects with CSA-CSR and stable congestive heart failure were randomised to 8 weeks treatment: oxygen 2 L/min through nasal prongs and concentrator or 8 weeks adaptive servo-ventilation (ASV) using a crossover design separated by a 3-week washout. Polysomnography, indices of sleep and breathing, shuttle walk distance, symptoms, urinary catecholamines, plasma brain natriuretic peptide (NT-BNP) and echocardiography were collected at baseline and completion of each arm.

Results: Ten subjects (age 64 ± 10 years, left ventricular ejection fraction (LVEF) 28 ± 10.5%, apnoea-hypopnoea index (AHI) 63 ± 30/h) were recruited. Seven completed the protocol (one died, one refused ASV, one was withdrawn after hospital admission). On therapy, an AHI of < 10/h was achieved in two out of seven using oxygen (29%), six of seven using ASV (86%) and six of seven with either (86%). Compliance with ASV: 5.2 ± 2 h/night (range 1.45-7.1 h/night). Median AHI on oxygen therapy: 13.4 /h (range 2.6-42.9/h), ASV, 1.4 /h (range 0.6-17.8/h, P = 0.03). LVEF was not changed by either therapy (oxygen 30.9% vs 30.9% P = 0.97, ASV 32.5% vs 35.0% P = 0.24). NT-BNP, urinary catecholamines, shuttle walk distance and symptoms were not significantly changed by either therapy.

Conclusion: CSA-CSR is reduced to a greater extent by ASV than oxygen therapy over 8 weeks but was not accepted long term. Neither treatment improved prognostic indices of heart failure or symptoms in the short term.