Combination pharmacological therapy targeting multiple mechanisms of sleep apnoea: a randomised controlled cross-over trial

Rationale Acetazolamide and atomoxetine-plus-oxybutynin (‘AtoOxy’) can improve obstructive sleep apnoea (OSA) by stabilising ventilatory control and improving dilator muscle responsiveness respectively. Given the different pathophysiological mechanisms targeted by each intervention, we tested whether AtoOxy-plus-acetazolamide would be more efficacious than AtoOxy alone. Methods In a multicentre randomised crossover trial, 19 patients with moderate-to-severe OSA received AtoOxy (80/5 mg), acetazolamide (500 mg), combined AtoOxy-plus-acetazolamide or placebo at bedtime for three nights (half doses on first night) with a 4-day washout between conditions. Outcomes were assessed at baseline and night 3 of each treatment period. Mixed model analysis compared the reduction in Apnoea-Hypopnoea Index (AHI) from baseline between AtoOxy-plus-acetazolamide and AtoOxy (primary outcome). Secondary outcomes included hypoxic burden and arousal index. Results Although AtoOxy lowered AHI by 49 (33, 62)% baseline (estimate (95% CI)) vs placebo, and acetazolamide lowered AHI by+34 (14, 50)% baseline vs placebo, AtoOxy-plus-acetazolamide was not superior to AtoOxy alone (difference: −2 (−18, 11)% baseline , primary outcome p=0.8). Likewise, the hypoxic burden was lowered with AtoOxy (+58 (37, 71)% baseline ) and acetazolamide (+37 (5, 58)% baseline ), but no added benefit versus AtoOxy occurred when combined (difference: −13 (−5, 39)% baseline ). Arousal index was also modestly reduced with each intervention (11% baseline –16% baseline ). Mechanistic analyses revealed that similar traits (ie, higher baseline compensation, lower loop gain) were associated with both AtoOxy and acetazolamide efficacy. Conclusions While AtoOxy halved AHI, and acetazolamide lowered AHI by a third, the combination of these leading experimental interventions provided no greater efficacy than AtoOxy alone. Failure of acetazolamide to further increase efficacy suggests overlapping physiological mechanisms. Trial registration number NCT03892772 .