Why the Next Generation of Sleep Apnea Drugs Targets Airway Muscles Instead of Body Weight

The sleep apnea drug pipeline now aims at airway muscles instead of body weight

The sleep apnea drug pipeline is the group of medicines in clinical development for obstructive sleep apnoea, and its centre of gravity has moved. For the past few years the story belonged to weight loss, and the evidence behind the weight-loss medicines is real for people whose apnoea is driven by obesity. The newest candidates aim somewhere else entirely.

They target the muscles that hold your throat open during sleep, and the feedback loop that decides how hard you breathe. Most adults with obstructive sleep apnoea do not have obesity, so a drug that only shrinks body fat leaves the majority behind.

A 2025 individual participant data meta-analysis in eClinicalMedicine (The Lancet Discovery Science) pooled 12,860 adults from four community cohorts, 7,222 of whom had OSA. Among those with OSA, 44.4% had overweight and 23.5% had a normal weight or were underweight. Normal weight OSA is not a rare exception.

Why your throat muscles let go the moment you fall asleep

Your upper airway contains no bone or cartilage. It stays open while you are awake because muscles actively pull it open, chiefly the genioglossus, the large muscle at the base of your tongue. These pharyngeal dilator muscles compensate constantly for a narrow throat.

When you fall asleep, the noradrenergic neurons that drive those muscles slow down. Their firing frequency falls progressively through NREM sleep and falls furthest in REM. Researchers describe this loss of noradrenergic drive as the dominant cause of sleep-related muscle relaxation in the throat.

That biology explains the shape of the leading drug candidate. A norepinephrine reuptake inhibitor restores part of the missing drive to genioglossus muscle activity. Researchers believe a second, muscarinic brake also suppresses the muscle in sleep, which is why the front-runner is a noradrenergic-antimuscarinic combination rather than either drug used alone.

How much that relaxation costs you depends on your anatomy. Clinicians measure it as upper airway collapsibility (Pcrit) and upper airway muscle responsiveness, meaning how strongly your dilator muscles recruit as the airway narrows. Retrognathia, narrow craniofacial anatomy, REM-related OSA and positional OSA can all produce a collapsible airway at a normal body weight.

Four endotypes decide whether a pill can work for you

An endotype is the underlying mechanism producing your apnoea, rather than the symptoms it creates. Sleep researchers group them on the PALM scale, covering Pcrit (anatomy), Arousal threshold, Loop gain and Muscle responsiveness, and this is the map that personalised OSA treatment is built on.

A 2019 review in the Journal of Clinical Medicine estimated how often each dominates. Anatomical defects are the main cause in about 46% of patients. Insufficient upper-airway muscle recruitment plays an important role in roughly 36%, high loop gain in 36%, and a low respiratory arousal threshold was found in 88%, 73% and 23% of patients with mild, moderate and severe OSA. The authors caution these estimates come from a limited single-laboratory cohort.

Most patients carry more than one endotype at once. These numbers help explain why an anatomy-dominant patient may respond poorly to a muscle-tone drug that helps someone with an identical apnea-hypopnea index (AHI).

Loop gain explained, and why an old European epilepsy medicine targets it

Loop gain measures how strongly your breathing control system reacts to a change in carbon dioxide. Picture an over-sensitive thermostat that blasts the heating, overshoots, then shuts off and lets the room go cold.

In high loop gain, a small disturbance triggers an oversized burst of ventilation. You blow off too much carbon dioxide, drive then undershoots, and with almost no drive reaching the dilator muscles the airway collapses. This is ventilatory control instability, a breathing-control problem rather than a throat problem.

Carbonic anhydrase inhibitors are used to damp that oscillation. By changing how the body handles carbon dioxide, they are thought to steady respiratory drive between breaths. Acetazolamide is the best-known member of that class; sulthiame is the candidate now tested in sleep apnoea trials.

The evidence is European. In the Phase 2 FLOW dose-finding trial, presented at the European Respiratory Society Congress in Vienna and published in The Lancet, 298 adults with untreated moderate-to-severe OSA were treated across 28 centres in Spain, France, Belgium, Germany and the Czech Republic for 15 weeks. As reported by Respiratory Therapy in 2025, AHI fell 17.8% at 100 mg, 34.8% at 200 mg and 39.9% at 300 mg, a clear dose-response, with mild to moderate side effects: pins and needles, headache, fatigue and nausea.

Sulthiame is not new. Synthesised in the 1950s, it is licensed for epilepsy in Germany, Austria, the Czech Republic, Hungary, Slovakia and Switzerland, and has never been registered in the United States. The loop-gain arm of the pipeline is a European repurposing story usually told from an American angle.

Does the new muscle-tone tablet still work if you are a normal weight?

On the published evidence, yes. AD109 combines aroxybutynin, an antimuscarinic, with atomoxetine, a norepinephrine reuptake inhibitor, in one once-nightly pill aimed at muscle responsiveness.

In the Phase 3 SynAIRgy trial of 646 adults across 69 centres, published in the American Journal of Respiratory and Critical Care Medicine in 2026, AD109 produced a 55.6% reduction in apnea-hypopnea index versus placebo at 26 weeks (p at or below 0.0001). Disease control, meaning an AHI below 5 events per hour, was reached by 22.3% of participants, 39.6% had AHI reductions of at least 50%, and hypoxic burden fell 60.5%.

The line that matters most to you sits in the same publication. Results were consistent in participants with and without obesity, spanning healthy weight, overweight and obesity. The earlier Phase 2 MARIPOSA trial pointed the same way, with a mean 47.1% reduction in AHI4 versus placebo, according to the 2025 Phase 3 design paper indexed in PubMed Central. That paper records a mean baseline BMI of 32.1 kg/m2 in the 660-participant LunAIRo trial and 32.3 kg/m2 in SynAIRgy, so that consistency comes from within-trial comparisons, not a separate lean-population study. See the Phase 3 numbers broken down in detail.

Now the part the headlines skip. In the same SynAIRgy Phase 3 trial, 21% of participants discontinued because of side effects: dry mouth, nausea, insomnia and urinary difficulty, according to the American Thoracic Society in 2026. That release also reported AHI falling about 44% on AD109 versus about 18% on placebo, and 18% achieving complete control.

What the sleep apnea drug pipeline actually means in Europe right now

No medicine is currently authorised in the European Union to treat the airway obstruction itself. Europe does have one EMA marketing authorisation tied to OSA that the United States lacks: pitolisant, authorised on 1 September 2021 after a CHMP opinion on 20 May 2021, held by a Paris-based company. It improves wakefulness and reduces excessive daytime sleepiness in adults with OSA whose sleepiness has not improved on, or who cannot tolerate, CPAP.

The EMA overview is explicit that this treats the sleepiness, not the apnoea. In the HAROSA III trial published in the Journal of Sleep Research in 2025, 361 patients were randomised, 242 to pitolisant and 119 to placebo. Pitolisant improved the Epworth Sleepiness Scale by a least-squares mean difference of -2.6 versus placebo (95% CI -3.4 to -1.8, p below 0.001). Solriamfetol sits in the same wake-promoting category, addressing residual symptoms rather than AHI.

There is also no European equivalent of the American weight-loss drug approval for sleep apnoea. In December 2024 the EMA concluded that tirzepatide does not need a separate obstructive sleep apnoea indication, judging that data from two studies in 469 adults with moderate-to-severe OSA and obesity already fell within the existing weight-management authorisation.

AD109's regulatory filing is United States only. The FDA, the American regulator, accepted the application with a target action date of 28 February 2027, and no EMA submission has been announced, so European availability will lag the American headlines.

One more European candidate needs a warning label. BAY2586116, a topical TASK-1/3 potassium channel antagonist delivered as a nasal spray, increases dilator muscle activity and reduces pharyngeal collapsibility. It reduced OSA severity only in physiological responders. That implies an endotyping step which is not part of routine European sleep-lab workup, so this class may reach clinics before the test that decides whether it works for you.

What to do while the sleep apnea drug pipeline matures

Start with a measurement, not a guess. You cannot choose between these mechanisms without knowing your own numbers, and the years before these medicines arrive still have to be slept through.

A nasopharyngeal airway stent belongs in that mechanical category, as a bridge rather than a pharmacological equivalent. Back2Sleep is a CE-certified Class I soft silicone intranasal stent from a French company in Paris that keeps the nasal airway open during sleep, in the same anatomical territory the nasal-spray candidates are trying to reach pharmacologically. It needs no prescription, no electricity, noise or tubing, and the starter kit holds four sizes for around EUR 39 with a 30-day money-back period.

Be equally clear about its limits. It does not increase genioglossus tone, does not lower loop gain, and does not treat central apnoea. It is intended for snoring and mild-to-moderate OSA only, never for severe disease, and it is not a CPAP substitute in severe apnoea. If high loop gain is your dominant trait, roughly 36% of patients, you are waiting on a carbonic anhydrase inhibitor rather than a stent.

Frequently Asked Questions