Rationale
Many patients with obstructive sleep apnea (OSA) are unable to tolerate long-term positive airway pressure (PAP) therapy, highlighting the need for alternative treatments. AD109 (investigational fixed-dose oral combination of aroxybutynin 2.5 mg/atomoxetine 75 mg) is designed to target neuromuscular dysfunction in OSA.
Objectives
To evaluate the efficacy/safety of AD109 over 6 months in a diverse OSA population unable to use PAP.
Methods
SynAIRgy enrolled adults with mild-to-severe OSA who were intolerant to or refused PAP therapy into a randomized, double-blind, placebo-controlled, 26-week parallel-arm trial of AD109 vs placebo across 69 centers. The primary efficacy endpoint was change from baseline to week 26 in apnea–hypopnea index (AHI). Key secondary endpoints were oxygen desaturation index (ODI), Patient-Reported Outcomes Measurement Information System (PROMIS)–Fatigue T-score, hypoxic burden (HB), PROMIS-Sleep Impairment T-score, and proportion of participants with ≥50% AHI reduction.
Measurements and Main Results
A total of 646 eligible participants (median age 58 years, 49.3% female, median body mass index 32.4 kg/m2) were randomized. Median baseline AHI was 19.6 events/hour with 35% mild, 42% moderate, and 23% severe OSA. At week 26, mean AHI treatment difference was −4.0 events/hour (95% CI, −6.4 to −1.6; P = .001), representing a model-estimated 44.1% vs 17.6% decrease from baseline (P <.0001). AD109 demonstrated improvements in ODI and HB at week 26 vs placebo; however, no statistically significant difference was observed for PROMIS-Fatigue. Overall, 21.2% of participants on AD109 and 3.1% on placebo discontinued therapy due to adverse events. The most common adverse events with AD109 were dry mouth, nausea, insomnia, and urinary hesitation, with no serious treatment-related adverse events.
Conclusions
AD109 significantly improved airway obstruction and oxygenation at 26 weeks across a broad range of patients unable to use PAP, suggesting that AD109 could become a potential treatment option for patients with OSA.
Rationale
Many patients with obstructive sleep apnea (OSA) are unable to tolerate long-term positive airway pressure (PAP) therapy, highlighting the need for alternative treatments. AD109 (investigational fixed-dose oral combination of aroxybutynin 2.5 mg/atomoxetine 75 mg) is designed to target neuromuscular dysfunction in OSA.
Objectives
To evaluate the efficacy/safety of AD109 over 6 months in a diverse OSA population unable to use PAP.
Methods
SynAIRgy enrolled adults with mild-to-severe OSA who were intolerant to or refused PAP therapy into a randomized, double-blind, placebo-controlled, 26-week parallel-arm trial of AD109 vs placebo across 69 centers. The primary efficacy endpoint was change from baseline to week 26 in apnea–hypopnea index (AHI). Key secondary endpoints were oxygen desaturation index (ODI), Patient-Reported Outcomes Measurement Information System (PROMIS)–Fatigue T-score, hypoxic burden (HB), PROMIS-Sleep Impairment T-score, and proportion of participants with ≥50% AHI reduction.
Measurements and Main Results
A total of 646 eligible participants (median age 58 years, 49.3% female, median body mass index 32.4 kg/m2) were randomized. Median baseline AHI was 19.6 events/hour with 35% mild, 42% moderate, and 23% severe OSA. At week 26, mean AHI treatment difference was −4.0 events/hour (95% CI, −6.4 to −1.6; P = .001), representing a model-estimated 44.1% vs 17.6% decrease from baseline (P <.0001). AD109 demonstrated improvements in ODI and HB at week 26 vs placebo; however, no statistically significant difference was observed for PROMIS-Fatigue. Overall, 21.2% of participants on AD109 and 3.1% on placebo discontinued therapy due to adverse events. The most common adverse events with AD109 were dry mouth, nausea, insomnia, and urinary hesitation, with no serious treatment-related adverse events.
Conclusions
AD109 significantly improved airway obstruction and oxygenation at 26 weeks across a broad range of patients unable to use PAP, suggesting that AD109 could become a potential treatment option for patients with OSA.
I don't think PAP/other treatments work for fatigue at all in OSA unless they actually significantly reduce inspiratory flow limitation (IFL) (apnea is not IFL; hypopnea is — though a lot of milder IFL will not meet criteria for hypopnea), so it doesn't surprise me at all that reducing the AHI by a mean of only 4 events per hour (from a starting median AHI of 19.6) with this drug has no impact on fatigue. However, it is encouraging they actually used a fatigue rating scale rather than just the Epworth Sleepiness Scale.
First of all, CPAP, as it is conventionally prescribed, does not actually seem to work that well for many OSA patients. In the European Sleep Apnea Database cohort of 4,853 CPAP-treated patients, Bonsignore et al. (2021) found that excessive daytime sleepiness (EDS, Epworth Sleepiness Scale >10) declined form 56% at baseline to 28.2% at follow-up (median 5 months), although longer treatment time made a difference: EDS was 40% at 0-3 month first follow-up visit and declined to 13–19% at 4 months - 2 years first follow-up visit. Incomplete resolution of apneas/hypopneas and/or poor CPAP adherence occurred with similar frequency in patients with and without persistent EDS, suggesting that residual symptoms are not simply a product of inadequate treatment hours or residual apnea/hypopnea events.
If you've been following along, you'll know that the Epworth Sleepiness Scale (ESS) actually measures a mix of objective sleepiness and fatigue, which are uncorrelated symptoms in OSA patients. I'll get into some reasons why I think CPAP may not work that well (at least on a shorter time scale) for many OSA patients with objective sleepiness in a later write-up, but that's not the type of OSA we are concerned with here when it comes to a potential connection to ME/CFS; we're concerned with the fatigue subtype of OSA (/UARS - there is no separate UARS), which is proposed to be driven by a physiological stress response to inspiratory flow limitation (IFL).
Which brings me to my next point: historically, CPAP titration protocols have been designed to eliminate apneas/hypopneas, but not necessarily IFL. There was one trial comparing a conventional titration protocol to an IFL-targeted titration protocol (see below), but considering ESS and Maintenance of Wakeful Test were the outcome measures -- and not fatigue/other symptoms relevant to this subtype of OSA -- I don't think the results are necessarily that informative, even though IFL-targeted titration did seem to result in better outcomes on some of the measures assessed.
If someone has the IFL-driven fatigue OSA subtype and they are titrated on CPAP to a pressure that only resolves apneas/hypopneas but not IFL, you would likely not expect them to feel significantly better; in fact, there's an argument that they might actually feel worse if they are now spending less time in apnea (complete cessation of airflow) and more time in IFL (and according to Dr. Gold's hypothesis of prolonged decreases in nasal pressure during IFL activating the limbic system, it's actually milder IFL that would be expected to activate the limbic system more than hypopnea [which is also a form of IFL] due to greater prolonged decreases in nasal pressure, as there is less airflow/negative nasal pressure during hypopnea than milder IFL). So CPAP potentially shifting people from hypopnea-dominant breathing to milder IFL-dominant breathing could also theoretically make things worse. You've also added the stress of CPAP (mask on your face, and potentially pressurized air itself -- especially in the case of CPAP over bilevel devices -- acting as a unique stressor too; see following thread for a discussion of that: Sleep-disordered breathing (UARS/OSA) and chronic insomnia).
The CPAP pressure baseline shift
Looking at Gold's Figure 5, the key thing to understand is that the y-axis represents airflow in ml/sec, with inspiration going downward (negative values) representing subatmospheric nasal pressure. This is at atmospheric baseline — no CPAP.
On CPAP, the entire pressure baseline is shifted upward by whatever the CPAP pressure is. So if someone is on 12 cmH₂O CPAP (which is the median optimal pressure in Parekh et al.), Inspiratory effort would need to generate a negative swing exceeding 12 cmH₂O just to bring nasal pressure to atmospheric, let alone below it.
So the honest answer to your direct question [is there still prolonged subatmospheric nasal pressure during IFL on CPAP?] is: probably not. IFL occurring at significantly elevated CPAP pressures is almost certainly not producing subatmospheric nasal pressure at the nares. The prolonged inspiratory duration and constrained flow plateau would still be present — the temporal signature of IFL — but the absolute nasal pressure environment is fundamentally different from atmospheric-baseline IFL.
Looking at the paper, the optimal CPAP pressure was a median of 12 cmH₂O, and suboptimal was 4 cmH₂O below that — so median 8 cmH₂O. Still well above atmospheric. And yet:
Sustained IFL (SIFL) at these pressures significantly increased K-complex density
K-complexes shifted from delta-dominant to alpha-dominant during SIFL periods
This shift explained 17% of additional PVT lapse variance, independent of AHI and arousal index
Crucially, there were no scorable cortical arousals and no oxygen desaturation during SIFL
This is a striking finding, because it suggests the CNS remains responsive to sustained flow limitation even under CPAP conditions where prolonged subatmospheric nasal pressure at the nares would be expected to be substantially reduced. The brain is registering something about flow-limited breathing that is arousal-promoting at the microarchitectural level, without any of the classical triggers. It should be noted that Parekh did not directly measure nasal pressure in the sense relevant to Gold's theory, so the inference about subatmospheric pressure is indirect rather than directly demonstrated.
What this means for the olfactory-limbic hypothesis
This raises an interesting question for Gold's framework. If the olfactory-limbic mechanism is specifically about prolonged subatmospheric nasal pressure as the sensory input, then Parekh's findings — CNS activation during IFL at CPAP pressures where that signal would be substantially attenuated — suggest either that the mechanism is broader than the subatmospheric pressure framing implies, or that parallel pathways also contribute:
The olfactory-limbic route — operative not only via prolonged subatmospheric nasal pressure but potentially via the distinctive temporal dynamics of IFL itself: the rapid early rise in inspiratory flow followed by a sustained plateau creates an airflow/pressure signature that is qualitatively different from normal breathing regardless of absolute pressure baseline, and could constitute a novel olfactory nerve stimulus even during CPAP-pressurized breathing. Gold's own Figure 5 emphasizes prolonged inspiratory duration and constrained flow contour — not just pressure amplitude — which may support a pattern-recognition interpretation of the olfactory signal more than a pure negative pressure intensity model. Whether Gold's proposed mechanism requires subatmospheric pressure specifically, or responds to abnormal flow dynamics more broadly, remains an open question worth clarifying with Dr. Gold directly.
Mechanoreceptor/respiratory effort signaling — the prolonged inspiratory effort itself, sensed via pulmonary stretch receptors or upper airway mechanoreceptors, which would operate regardless of CPAP pressure level.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
