Atomoxetine plus oxybutynin for obstructive sleep apnoea: a randomised-controlled trial with detailed pathophysiology.

Abstract

Obstructive sleep apnoea (OSA) is a prevalent disorder with major health consequences for which there is no approved pharmacological therapy targeting upper airway muscle pathophysiology. A combination of a noradrenergic (atomoxetine) and antimuscarinic (oxybutynin) (AtoOxy) may meaningfully reduce apnoea-hypopnoea index (AHI), but further repeated-dose intervention data are needed. 58 participants with moderate-to-severe OSA were randomised to receive AtoOxy (80/5 mg) and placebo in a 1-month cross-over study. Primary analysis quantified the effect of AtoOxy versus placebo on AHI (percent change from baseline). Individual pathophysiology was characterised at baseline in a stand-alone gold-standard physiology study to determine whether OSA traits (collapsibility per 'Pcrit'; arousal threshold and muscle effectiveness per intraoesophageal catheter; loop gain) modified AtoOxy effectiveness. AHI was lowered by -23.8 (-35.2, -10.6)%_baseline (estimate (95% CI)) with AtoOxy versus by -11.7 (-24.3, 2.9)%_baseline with placebo; the treatment difference of -12.1 (-22.4, -0.5)%_baseline was significant (p=0.041). Only arousal threshold was a significant modifier of response: AtoOxy treatment effect was greater in high versus low arousal threshold (-25 vs +3 %_baseline treatment difference from placebo). Point-estimate treatment differences were observed within low but not high muscle effectiveness (-20 vs -4 %_baseline) and mild but not severe collapsibility (-17 vs -7 %_baseline). Sensitivity analysis suggested a twofold greater treatment difference using AHI₄ (4% hypopnoea criteria). AtoOxy reduced AHI over 1 month; although the average response was not clinically meaningful, greater improvements were observed in participants with a higher arousal threshold, lower muscle effectiveness and milder collapsibility, which could enable targeted pharmacological intervention.