A Pill for Sleep Apnea Cut Breathing Interruptions by 44% in a Phase 3 Trial

Millions of people with obstructive sleep apnea cannot or will not use CPAP, the standard treatment. A pill may give them another option.

In the Phase 3 SynAIRgy trial, AD109 — an oral combination of aroxybutynin and atomoxetine — reduced breathing interruptions by 44%, against 18% for placebo. Results were published in the American Journal of Respiratory and Critical Care Medicine.

What obstructive sleep apnea actually is

The name describes the mechanism precisely. During sleep, muscle tone throughout the body falls, including in the muscles holding the upper airway open. In susceptible people — because of anatomy, weight, or the degree to which tone drops — the airway collapses. Breathing stops until oxygen falls and carbon dioxide rises far enough to trigger a brief arousal, the airway reopens, and the cycle repeats.

This can happen dozens of times an hour, all night. The consequences are not confined to tiredness: repeated oxygen desaturation and arousal are associated with hypertension, cardiovascular disease, metabolic dysfunction and impaired daytime function.

Why CPAP is both excellent and inadequate

Continuous positive airway pressure works by brute force — a mask delivering pressurised air that physically splints the airway open. When used, it is highly effective, which is why it has been standard for decades.

The problem is not efficacy but use. A large proportion of patients prescribed CPAP do not use it adequately, and many abandon it entirely. Sleeping with a mask strapped to your face, tethered to a machine, is genuinely difficult for many people, and the dropout is substantial and well documented. A treatment that works perfectly and is not used delivers nothing.

What AD109 does differently

AD109 targets the neuromuscular control of the airway rather than mechanically forcing it open. Its two components act on the neural signalling that sustains muscle tone in the upper airway, aiming to keep the dilating muscles — principally those controlling the tongue — sufficiently active during sleep to resist collapse.

“An oral pill targeting neuromuscular drivers of airway collapse could help address this gap and broaden effective options,” said the University of Pittsburgh’s Dr Patrick Strollo.

That is a conceptually different intervention from anything currently available in pill form, and if approved it would be the first drug treating the mechanics of the condition.

The trial

SynAIRgy enrolled 646 adults with mild-to-severe sleep apnea who had refused or could not tolerate CPAP, across 69 sites in the US and Canada, over six months.

The population is the right one. Testing a drug in patients who have already declined the effective standard treatment answers the clinically relevant question — what to offer people who will otherwise go untreated — rather than asking whether a pill can beat CPAP in patients happy to use it.

Beyond the headline reduction, more than 40% of patients improved to a less severe category and 18% achieved complete disease control.

How to read a 44% reduction

The placebo arm improved 18%, so the drug’s incremental effect is smaller than the headline suggests. That placebo response is typical in sleep apnea trials, where measurement involves a sleep study and night-to-night variability is considerable.

More importantly, a 44% reduction in breathing interruptions is not the same as resolution. Someone with severe apnea reduced by 44% still has significant disease. That is why the categorical measures matter more: 40% moving to a less severe category and 18% achieving complete control describe how many people were meaningfully helped, rather than an average across everyone.

Tolerability

Side effects — dry mouth, nausea, insomnia and urinary difficulty — led about 21% to discontinue.

That figure deserves attention given the drug’s entire rationale. AD109 exists because patients will not tolerate CPAP; a one-in-five discontinuation rate over six months means it inherits an adherence problem of its own, if a different one. Insomnia is a particularly awkward adverse effect in a therapy for a sleep disorder.

Where it fits among the alternatives

AD109 would not arrive into an empty field. Surgical and implanted options exist for selected patients, including devices that stimulate the nerve controlling tongue position during sleep — effective but invasive and expensive. Dental appliances help milder cases.

The most significant recent development is pharmacological: a GLP-1 receptor agonist has been approved for moderate-to-severe obstructive sleep apnea in adults with obesity, working by reducing the weight that contributes to airway collapse. That addresses a cause rather than the mechanism, and applies only where obesity is the driver.

AD109 targets a different mechanism entirely and is not restricted by weight, so the two would be complementary rather than competing. Experts consistently frame it as an alternative rather than a replacement for CPAP.

Where it stands

AD109 has FDA Fast Track designation, a New Drug Application has been submitted, and a decision is expected in early 2027.

The open questions are durability beyond six months, whether the benefit translates into the cardiovascular and metabolic outcomes that make sleep apnea worth treating, and how the discontinuation rate behaves outside a trial. None of those will be answered before approval. This summarises trial results and is not medical advice.