Reimagining obstructive sleep apnoea: a new era for sleep medicine

As The Lancet Respiratory Medicine marks its 10-year anniversary, the spotlight on obstructive sleep apnoea (OSA) has never been brighter. This past decade has reshaped our understanding of OSA pathophysiology and incited questions pertaining to the role of OSA therapy. As our understanding evolves, pressing questions emerge: do all OSA patients need treatment and, if so, which individuals require it most urgently? Moreover, how do different OSA treatments influence health outcomes? Researchers have made important insights into the effects of OSA therapy on patient-centred and cardiovascular outcomes, with substantial advancements in OSA classification using innovative physiological metrics and the incorporation of machinelearning algorithms and omics, paving the future for precision health in OSA.

OSA affects nearly 1 billion individuals worldwide and is characterised by repetitive upper-airway collapse during sleep, resulting in intermittent hypoxaemia and sleep fragmentation. OSA is associated with various cardiovascular diseases and conditions, including hypertension, atrial fibrillation, heart failure, and stroke. Currently, OSA is defined using the apnoea–hypopnoea index (AHI), which captures the frequency of respiratory events per hour of sleep. Continuous positive airway pressure (CPAP) is the cornerstone of OSA therapy and is effective at improving daytime sleepiness. However, multiple randomised controlled trials over the past decade have not shown a benefit of CPAP therapy in the prevention of composite cardiovascular events among non-sleepy patients with moderate-to-severe OSA.^1,2

In the past decade, clinical sleep medicine has transitioned from the use of exhaustive overnight polysomnography to streamlined one-night home studies. This transformational shift was mainly driven by the perceived importance of the AHI as the sole metric of OSA severity. However, overemphasising this index neglects crucial physiological data, hindering accurate prediction of cardiovascular risk and a holistic understanding of patient health—including the intricate sleep-related interplay between the brain, heart, and lungs. In particular, the AHI fails to capture the array of repetitive physiological disruptions, including the variability in respiratory event duration, the depth and duration of oxygen desaturation, the pulse-rate response to respiratory events, and the ventilatory burden, among others. Several studies have shown that OSA-specific hypoxic and ventilatory burden predict cardiovascular disease and all-cause mortality.^3,4 Additionally, individuals with OSA and an elevated pulse-rate response to respiratory events are at increased risk of cardiovascular morbidity and mortality.⁵ Short respiratory event duration, a marker of low arousal threshold, also predicts all-cause mortality.⁶ This robust evidence supports the use of novel physiological metrics that are not fully accounted for by the AHI. The AHI also fails to capture symptom-based OSA phenotypes. Using machine-learning techniques, a 2019 study found that the excessively sleepy OSA subtype is linked to an increased prevalence of heart failure and an increased risk of incident cardiovascular disease and coronary disease.⁷

Research over the past two years has also cast doubt on the commonly held assumption that CPAP therapy is entirely benign. One study⁸ showed an increase in circulating concentrations of angiopoeitin-2, a potent driver of vascular inflammation, after CPAP. Increased concentrations of this protein might reflect mechanical effects of CPAP on the lung, with alveolar distension and resultant endothelial cell dysfunction. Another study⁹ used proteomics and machine learning to identify three unique clusters of patients with OSA on the basis of inflammatory protein expression. Exploratory analysis found that the clusters responded differently to CPAP, with an increase in inflammatory protein expression in the low-inflammatory cluster and a decrease in the high-inflammatory cluster.

Given the aforementioned work, it is clear that we have entered an era of research that will not only pinpoint high-risk OSA subgroups but also introduce novel OSA metrics that comprehensively encompass symptoms and physiological data. These OSA phenotypes will serve as crucial prognostic biomarkers for cardiovascular disease risk stratification and will guide patient selection in future trials. Future research should explore how OSA therapies affect these distinct subgroups while maintaining patient safety as a paramount consideration.

Finally, the past decade has also brought substantial advancements in OSA therapies (panel). Hypoglossal nerve stimulation of the genioglossus muscle has been shown to significantly improve OSA severity among a subgroup of patients with moderate-to-severe OSA.¹² Additional surgical methods, such as ansa cervicalis stimulation of the sternothyroid muscle, have also shown promise and are currently being tested.¹³ Other novel therapeutic approaches target obesity, one of the strongest risk factors for OSA. These approaches include the use of glucagon-like peptide-1 receptor agonists, with an ongoing placebo-controlled phase 3 trial¹¹ assessing efficacy and safety in patients with obesity who have moderate-to-severe OSA. Another exciting direction involves the use of novel oral pharmacotherapies. At the forefront is a combination of the selective norepinephrine-reuptake inhibitor (atomoxetine) and the antimuscarinic agent (aroxybutynin), which activates upper-airway dilator muscles by targeting key neurological pathways. This combination has shown promising results in a phase 2 trial.¹⁰

Panel: Exploring innovative approaches to OSA treatment: ongoing and upcoming clinical trials*.

Pharmacological therapy

• The MARIPOSA trial (NCT05071612) showed promising results with the combination of a selective norepinephrine-reuptake inhibitor (atomoxetine) and an antimuscarinic agent (aroxybutynin) versus placebo for OSA in a phase 2 multicentre RCT, with phase 3 approved by the US Food & Drug Administration.¹⁰

• The ACE-Of-HEARTs trial (NCT05616260) is a double-blind phase 2 RCT to assess the efficacy of the diuretic acetazolamide on OSA severity and cardiovascular health in young adults (age 18–50 years) with moderate-to-severe OSA and hypertension at 2 weeks. Similarly, the PANACEA trial (NCT05804084) will evaluate the efficacy of acetazolamide on the AHI and neurocognitive outcomes in adults with OSA at 4 weeks.

Hypoglossal and transoral neurostimulation

• The OSPREY trial (NCT04950894) is a confirmatory, 13-month, multicentre, open-label RCT of the aura6000(R) system for targeted hypoglossal nerve stimulation (a simpler technology than hypoglossal nerve stimulation, with no sensor and an easier, proximal electrode implantation) in moderate-to-severe OSA, evaluating the difference in AHI response rates between the treatment and control groups at 7 months. This study is a follow-up to the THN3 study, which established a favourable safety profile and good AHI response rates, but with improved functionality of the system design.

• The ELMO study (NCT05252156) is an open-label RCT assessing the efficacy of intraoral neuromuscular electrical stimulation for moderate OSA using the eXciteOSA device, currently approved by the US Food & Drug Administration for snoring and mild OSA. The primary endpoint is the difference in AHI between control and therapy groups.

Targeting obesity

• The SURMOUNT-OSA trial (NCT05412004) is a phase 3 RCT investigating the efficacy and safety of the glucagon-like peptide-1 receptor agonist tirzepatide for the treatment of moderate-to-severe OSA in patients with obesity.¹¹ The primary endpoint is the difference in AHI response between the intervention and placebo groups at 52 weeks, among patients with OSA on CPAP and among those not on therapy.

Multimodal and other PAP-based OSA therapies

• The ESSENTIAL trial (NCT05988385) is 5-year, multicentre RCT recruiting cognitively healthy adults with moderate-to-severe OSA treated with any combination of multimodal therapy (eg, CPAP, dental sleep apnoea device, or positional therapy) to assess whether therapy results in an improvement of cognitive function and a change in biomarkers of Alzheimer’s disease.

• The ADVENT-HF trial (NCT01128816) is a multicentre RCT assessing the effects of adaptive servo-ventilation on survival and frequency of hospital admissions in patients with heart failure and moderate-to-severe sleep apnoea. The study completed recruitment in 2022.

AHI=apnoea–hypopnoea index. CPAP=continuous positive airway pressure. OSA=obstructive sleep apnoea. RCT=randomised controlled trial. *Articles and trials were selected on the basis of their relevance to the content of the current Comment.

In summary, the past 10 years have marked a pivotal era in sleep research. We’re now critically re-evaluating the need for universal OSA treatment, exploring OSA subgroups in terms of pathophysiology and symptoms, and recognising that CPAP might not be suitable for all. Advanced analytical techniques paired with translational research can provide profound insights into various OSA endophenotypes and guide personalised therapy in OSA, akin to asthma and cancer therapeutics. This shift away from the one-size-fits-all prototype acknowledges the heterogeneity of treatment effects on various health outcomes in OSA. Further, the sole reliance on the AHI is likely to be replaced by more tailored and informative measures in the coming years. Yet our journey is far from over. Integrating factors such as sleep duration and quality and circadian rhythms into future studies will be crucial to enhance our understanding of OSA and its broader implications. Moreover, uncovering existing disparities in sleep health and their effects on overall health outcomes is paramount. Addressing these dimensions ensures a comprehensive approach to personalised sleep medicine, enhancing patient care and outcomes. The evolution in OSA research not only reflects the groundbreaking work of the past decade but also foreshadows an exciting era of research and therapy focused on individualised patient care.