Background:Hypoglossal nerve stimulation (HGNS) is an established surgical therapy for obstructive sleep apnea (OSA), yet treatment response is variable and appears to be influenced by both body mass index (BMI) and pattern of upper airway collapse. Whether excess body weight differentially affects HGNS efficacy across pharyngeal collapse patterns remains unclear. Methods:We combined data from two independent HGNS cohorts (n=760) to examine whether the association between BMI and HGNS efficacy depends on pharyngeal collapse pattern identified by drug-induced sleep endoscopy. Collapse was categorized as predominantly laterally directed or anteroposterior (AP). Multivariable regression models assessed HGNS efficacy-quantified primarily as percent reduction in apnea-hypopnea index (AHI) on titration polysomnography and secondarily as treatment success (≥50% AHI reduction to <15 events/h)-while testing for effect modification by collapse pattern and adjusting for baseline AHI, partial collapse, surgical center, follow-up sleep study type, and prior or concomitant pharyngeal surgery. Results:Increasing BMI was associated with substantially lower HGNS efficacy among patients with laterally directed collapse, whereas the relationship between BMI and efficacy was attenuated in those with AP collapse. Specifically, each 5 kg/m2 increase in BMI was associated with a -19.7% (95% CI: -33.2 to -6.2) greater reduction in efficacy in lateral collapse compared with -3.8% (-8.0 to 0.36) in AP collapse (interaction p=0.027). Higher BMI also corresponded to reduced odds of treatment success in lateral collapse (odds ratio 4.4 [95% CI: 1.4-14.3] per 5 kg/m2), with no significant association observed in AP collapse (1.1 [0.75-1.5]; interaction p=0.023). Conclusions:The influence of BMI on HGNS treatment response differs meaningfully by pharyngeal collapse phenotype. Incorporating collapse pattern into BMI-based eligibility criteria may improve patient selection and optimize HGNS outcomes.

Objective:Several endotypes contribute to the development of Obstructive Sleep Apnea (OSA). However, efforts to measure these endotypes have been challenging. In this paper, we propose a new method that overcomes some of these challenges. Methods:To test the feasibility of this new method, data from the Sleep Heart Health Study (SHHS) were analyzed and two oxygen-based endotypes were identified and plotted on a graphical model: the steady-state SpO2 and the SpO2 arousal threshold. The first is the oxygen saturation that would occur during sleep if there were no arousals, and it is a measure of upper airway collapsibility (a more collapsible airway produces a lower SpO2). The latter is the oxygen saturation that triggers arousals. These endotypes were validated by assessing their ability to detect positional and state-related changes in airway collapsibility and arousal threshold. Results:The study showed that it was feasible to measure oxygen-based endotypes in 95% of SHHS participants. As expected, steady-state SpO2 was lower during supine vs. non-supine sleep, as well as during REM vs. NREM sleep. Also, the SpO2 arousal threshold was similar between supine and non-supine sleep. However, SpO2 arousal threshold was not lower in REM sleep vs. NREM sleep. Therefore, in 3 of the 4 conditions, the oxygen-based endotypes moved in the expected direction due to positional or sleep state changes. Conclusion:Although further validation experiments are required, this study indicates that OSA endotyping using the pulse oximetry signal is feasible. The oxygen-based endotypes could be used to aid therapeutic decision making.

BACKGROUND:Obstructive sleep apnea (OSA) is a heterogenous disease characterized by several endo-phenotypes (mechanistic traits and physiological severity metrics) and variable associations with excessive daytime sleepiness (EDS). Examining the relationship of EDS with features of OSA may identify physiological processes that drive increased risk for adverse health outcomes in patients with OSA and EDS. RESEARCH QUESTION:Do variations in physiological endo-phenotypes identified from polysomnography associate with EDS in a general population, and do such associations vary by sex? STUDY DESIGN AND METHODS:Participants in the Multi-ethnic Study of Atherosclerosis (MESA) Exam 5 Sleep Study underwent polysomnography, actigraphy, and questionnaire assessment. Among participants with an apnea-hypopnea index > 5 events/hour (n = 1783; mean, age 68.7 ± 9.1; 52% female), we examined associations between OSA endo-phenotypes including physiological severity measures (hypoxic burden, arousal intensity, event duration, heart rate response [Δ HR]), and endotypes (mechanistic traits: collapsibility, compensation, loop gain, and arousal threshold) with EDS (Epworth Sleepiness Scale, ESS > 10). Poisson regression models with robust variance were used, adjusting for demographics, smoking, and other factors. RESULTS:EDS was found in 14% of the sample. An interquartile increase in hypoxic burden was associated with a 28% higher prevalence ratio (PR) of EDS (PR = 1.28; 95% CI: 1.10, 1.51) in models adjusted for demographic factors and smoking. An increased adjusted PR of EDS was also associated with higher loop gain (1.26; 1.12, 1.42) and higher ΔHR (1.32; 1.01, 1.73). In sex-stratified analyses, EDS was associated with shorter respiratory event duration and higher compensation in females, and with hypoxic burden and high ΔHR in males (P value for sex interaction < .05 for event duration). Associations did not materially change with adjustment for short sleep duration. INTERPRETATION:EDS was associated with several OSA endo-phenotypes. EDS was more strongly associated with shorter event duration in females, with trends for stronger associations for hypoxic burden and ΔHR in males. The physiological correlates of EDS in people with OSA may contribute to differences in OSA outcomes across symptom groups.

PurposeTo characterize changes in ventilation and inspiratory airflow associated with visual evidence of epiglottis-related obstruction during drug-induced sleep endoscopy (DISE) with positive airway pressure (PAP) administration among adults with obstructive sleep apnea (OSA).MethodsThis was a case-control study of adults with OSA undergoing DISE with measurement of airflow and mask pressure, prior to and during PAP administration. Cases were defined by visual evidence of partial or complete epiglottis-related obstruction, whether prior to and/or with PAP. Each case was matched to a control with an identical visual obstruction pattern (other than for the epiglottis) prior to PAP (if available), with attempted matching on age and BMI. For each breath, ventilation and mean inspiratory airflow were calculated, along with the endoscopic degree of visual epiglottis-related obstruction.ResultsFrom 150 study participants, 26 cases with 2050 breaths and 14 matched controls with 1353 breaths were selected. With increasing PAP level and mask pressure, 19 cases (13% of total study participants) had some decrease in ventilation or mean inspiratory airflow. Increasing mask pressure was associated with increased ventilation and mean inspiratory airflow in controls at all pressures and in cases before the onset of visual epiglottis-related obstruction. Among cases, once visual epiglottis-related obstruction emerged, increasing mask pressure was associated with decreasing ventilation and mean inspiratory airflow relative to the pattern in controls and cases at lesser mask pressures.ConclusionVisual epiglottis-related obstruction during DISE with PAP occurs in some individuals that limits increases in ventilation and mean inspiratory airflow normally seen with increases in PAP levels.

BACKGROUND AND AIMS:Randomized trials of continuous positive airway pressure (CPAP) treatment for obstructive sleep apnoea (OSA) in patients with cardiovascular disease have not detected reduced risk of major adverse cardiovascular and cerebrovascular events (MACCEs). This study tested whether the cardiovascular benefit of CPAP occurs preferentially in high-risk OSA, characterized by greater OSA-related heart rate acceleration or hypoxaemia. METHODS:In a post hoc analysis of pooled Randomized Intervention with Continuous Positive Airway Pressure in Coronary Artery Disease and Obstructive Sleep Apnoea, Impact of Continuous Positive Airway Pressure on Patients with Acute Coronary Syndrome and Nonsleepy Obstructive Sleep Apnoea, and Sleep Apnoea Cardiovascular Endpoints Study randomized trials; outcomes were stratified by high-risk OSA status, defined by heart rate response following OSA respiratory events >9.4 b.p.m. (third tertile) or oxygen desaturation area under baseline (hypoxic burden) > 87.1% min/h (third tertile). Cox mixed models quantified the CPAP treatment effect on MACCE (including cardiovascular mortality, myocardial infarction, and stroke) within high-risk OSA and the difference vs low-risk status (primary test). Secondary analyses examined participants without excessive sleepiness (Epworth <11 points) or without increased blood pressure (systolic/diastolic <140/90 mmHg). RESULTS:In 3549 participants, 16.6% and 16.3% reached the MACCE endpoint with CPAP (n = 1778) and usual care (n = 1771), respectively. The CPAP treatment effect was greater in participants with vs without high-risk OSA [interaction hazard ratio (iHR) .69, 95% confidence interval (CI) .50-.95, Pinteraction = .024; Nhigh-risk = 1832]. The differential effect was stronger in those without excessive sleepiness (iHR .59, 95% CI .41-.84; Nhigh-risk = 1509), or without increased blood pressure (iHR .54, 95% CI .36-.81; Nhigh-risk = 1244). Continuous positive airway pressure benefits in high-risk OSA were observed alongside harm in low-risk OSA. CONCLUSIONS:Continuous positive airway pressure preferentially improves cardiovascular outcomes in high-risk OSA, while harm in low-risk OSA may counteract this effect. These findings provide a pathway to identify patients likely to benefit.

Background and objective: The study aims to determine whether respiratory inductance plethysmography (RIP) signals can be used to quantify changes in ventilation and provide advanced obstructive sleep apnea (OSA) severity metrics. This approach seeks to address limitations in current airflow-based OSA measures, particularly those relying on nasal pressure, which may be compromised by oral breathing. Methods: Adult patients with OSA (N = 89, 68Male:21Female) completed in-laboratory polysomnography (PSG) allowing for RIP-based ventilation estimates to be compared against a gold standard oronasal-pneumotach (normalized ventilation %(eupnea)). Concordance was assessed on three levels: 1) individual breath ventilation, 2) individual respiratory event depth (percentage reduction in ventilation from local average), and 3) patient-specific OSA severity in terms of average event depth and ventilatory burden (average event depth x average event duration x event rate). To address overestimation of RIP ventilation during obstruction, we developed and applied a calibration and linearization method ("RIP correction"). Concordance analysis evaluated median bias for both small (<70%(eupnea)) and large breaths (>130%(eupnea)), along with bias and intraclass correlation coefficient (ICC) calculation for events and patient-specific measures. Results: For individual breaths (N = 495,631), RIP correction reduced overestimation bias for small breaths from 12 to 2%(eupnea). For individual events (N = 34,497), RIP correction reduced mean bias for event depth estimates from 9 to 1%(eupnea). For patient-specific analysis underestimation of average event depth was attenuated from 9 to 4%(eupnea) and for ventilatory burden, from 275 to 117%(eupnea) min/hr. Additionally, RIP correction improved ICC for event depth and patient-level traits. Conclusion: RIP signals, with appropriate processing, enable quantification of advanced ventilation-based OSA metrics, addressing concerns that airflow-based measures may be affected by breathing route.

Residual sleep apnea occurs in ~20% of patients (defined by a residual apnea-hypopnea index, (rAHI) ≥10 events/hour) using positive airway pressure (PAP) therapy. High loop gain (LG) contributes to obstructive sleep apnea (OSA) by a waxing and waning neural input to the dilator muscles of the upper airway resulting in episodic airway obstruction. We hypothesized that high LG is associated with increased risk of high rAHI among people using PAP. We analyzed data from two completed randomized controlled trials of PAP therapy (APPLES and RICCADSA). The APPLES study evaluated the effects of fixed PAP therapy on neurocognition in adults with OSA. The RICCADSA study assessed the effects of autotitrating PAP on cardiovascular outcomes in adults with coronary artery disease and OSA. Our primary outcome was rAHI ≥10 events/hour evaluated by polysomnography on PAP at 2 months (APPLES) or by PAP device downloads at 3 months (RICCADSA). We measured LG from baseline polysomnography using a validated method (Terrill et al., 2015). We performed logistic regression in the discovery sample (APPLES) and then the validation sample (RICCADSA) with LG as the exposure (highest vs. other quartiles). We then adjusted for baseline AHI, age, sex, body mass index (BMI), and pharyngeal collapsibility. Seventeen percent of the discovery (APPLES) sample (n=448) and 14% of the validation (RICCADSA) sample (n=185) exhibited rAHI ≥10 events/hour while on PAP. In unadjusted analysis, high LG was associated with an increased risk of ≥10 residual events/hour in both discovery and validation samples: odds ratio (OR) 3.1 [1.9, 5.2] and 3.2 [1.3, 7.4] respectively. This risk remained significantly increased after adjustment for baseline AHI, age, sex, BMI and pharyngeal collapsibility in APPLES (2.2 [1.2, 3.8]) and RICCADSA (3.5 [1.4, 8.7]). A sensitivity analysis that adjusted for central events at baseline did not meaningfully alter the results. High LG on baseline polysomnogram is associated with an increased risk of residual sleep apnea while on PAP therapy (rAHI ≥ 10 events/hour). Identifying patients with this characteristic at start of PAP therapy may help select those who may benefit from adjunctive therapies (e.g., acetazolamide or oxygen) if residual sleep apnea occurs. 1K23HL159259

Hypoglossal nerve stimulation (HGNS) is an effective alternative to continuous positive airway pressure (CPAP) for obstructive sleep apnea. Although adherence to HGNS is generally high, approximately 15-27% of patients exhibit suboptimal adherence, limiting effectiveness. Adherence is particularly important for HGNS, as it involves a surgical procedure with inherent risks and significant costs. Low arousal threshold has been identified as a predictor of non-adherence to CPAP. Similarly, this study tests the hypothesis that a low arousal threshold is associated with decreased HGNS usage. We performed a secondary analysis of the Stimulation Therapy for Apnea Reduction (STAR) Trial. Arousal threshold and pharyngeal collapsibility (i.e., endotype traits) were estimated from baseline polysomnography studies. HGNS usage was calculated from cumulative hours of stimulation pulse duration recorded by the device and downloaded at the 12-month follow-up. Since this value reflects the hours of active stimulation during inspiration, it was divided by 40%, assuming a respiratory duty cycle of 40%, to estimate total HGNS on time (i.e., usage). To test our hypothesis, we used multiple linear regression to quantify the association between arousal threshold and HGNS usage, adjusting for age, sex, neck circumference, and total sleep time. Sensitivity analysis evaluated the effect of adjusting for collapsibility. The mediating effect of HGNS efficacy (%reduction in AHI from baseline) was also studied. HGNS usage data was available from 64 patients at the 12-month follow-up, among whom 50 had a baseline PSG to compute the endotype traits (median[IQR] Age: 53[45,60] years, 43 men, BMI: 27.8[26.0,30.0] kg/m2, AHI: 33.8[25.1,40.0], Usage: 6.2[4.2,8.4] hours). As hypothesized, a lower arousal threshold by 1SD (25.8 %eupneic ventilation), was associated with >1 hour less of HGNS usage (Beta [95%CI]: 1.1[0.17,2.0] hours, p=0.011). Including collapsibility in the model increased the effect of arousal threshold on usage (Beta [95%CI]: 1.2[0.26,2.2] hours). Including HGNS efficacy in the model did not mediate the effect of arousal threshold on usage (1.1[0.17,1.9]). Patients with a lower arousal threshold—those who wake from sleep more easily—show significantly reduced HGNS usage, an effect that is independent of HGNS efficacy. These findings uncover a critical mechanism driving suboptimal adherence to HGNS therapy.

Rationale: A major challenge for the development of an effective and tolerable pharmacological intervention for obstructive sleep apnea (OSA) is that noradrenergic muscle stimulants like atomoxetine have wake-promoting properties. Pimavanserin, a promising serotonin 2A receptor antagonist, may increase arousal threshold and help reduce OSA severity. With this randomized, crossover, double-blind clinical trial, we tested the effect of this drug combination on apnea-hypopnea index (AHI; primary outcome), arousal index and nadir oxygen saturation (SpO2; secondary outcomes). Methods: Following baseline polysomnography, 18 OSA participants (AHI>15events/h) took pimavanserin-plus-atomoxetine (34/80mg; 34/40mg for the first 3 days) or placebo for one-week; follow-up polysomnography was performed to provide study outcomes. Safety outcomes (including corrected QT (QTc) as both drugs can potentially increase QTc length), subjective sleep quality, and flow-estimated endotypes (using oronasal pneumotachograph flow) were also explored. Results: The atomoxetine-plus-pimavanserin combination reduced AHI by 42 [95%CI: 18, 60] % vs. placebo, meeting the primary outcome (P<0.001). Absolute AHI reduction was 16.9 [8.1, 23.6] events/h greater than placebo. Nadir SpO2 and arousal index were also improved, by 5.0 [1, 8] % and 10.9 [2.4, 18.1] events/h vs. placebo. Overnight heart rate was increased (+4.8 [1.5, 8.1]), but no change in QTc, subjective sleep quality or next-morning vital signs was evident. There was no increased risk for side effects on the combination vs. placebo. Treatment vs. placebo improved pharyngeal collapsibility (+7.9 [1.6, 14.1]%VEUPNEA), reduced loop gain by 20% (0.15 [-0.23, -0.07]), and did not reduce the arousal threshold. Individuals with less-severe collapsibility (greater VPASSIVE) at baseline experienced a more favorable AHI reduction on atomoxetine-pimavanserin compared to placebo (additional 25.0 [-11.5, 48.7] % for each 1SD increase in VPASSIVE). Conclusions: The combination of atomoxetine and pimavanserin significantly reduced OSA severity, lowering arousal index and overnight hypoxemia. Compared to placebo, the combination did not raise the risk of many adverse effects seen in similar trials (e.g., insomnia, headache, tachycardia), except for a mild heart rate increase. Pimavanserin shows promise to be associated to atomoxetine for treating OSA and merits further, larger trials.

BACKGROUND AND OBJECTIVES:Patient compliance with continuous positive airway pressure (CPAP) is similar using manual-titrated pressure compared to auto-titration, although auto-titration pressures are usually 2-5 cmH2O higher than manual pressure, indicating that CPAP moderately higher than the optimal pressure will not necessarily impair compliance. We try to find the tolerable highest CPAP which does not increase respiratory effort based on changes in lung volume, diaphragm electromyography (EMG) and breathing sensations in healthy volunteers and OSA patients to simplify pressure titration. METHODS:Part 1, 12 healthy subjects and 16 OSA patients were enrolled in the measurement of expiratory reserve volume, diaphragm EMG, and expiratory muscle EMG at different CPAP levels. Breathing difficulty during different CPAP levels was assessed using a customised questionnaire in 35 healthy subjects and 33 OSA patients. Part 2, a two-night randomised crossover double-blind trial using the tolerable highest CPAP (10 cmH2O) based on the results derived from Part 1 and the manually titrated pressure was performed in 25 OSA patients. RESULTS:End expiratory lung volume increased significantly with increasing CPAP. In general, diaphragm EMG changed little when CPAP ≤ 10 cmH2O. Expiratory muscle activity appeared when CPAP > 12 cmH2O. There was no significant difference in subjective sensation of breathing difficulty with CPAP ≤ 10 cmH2O. Sleep structure, AHI, and patient preference with 10 cmH2O CPAP were not different from those under titrated pressure. CONCLUSIONS:This study suggests that most patients with moderate to severe OSA can be effectively treated with CPAP at an initial pressure of 10 cmH2O without pressure titration. TRIAL REGISTRATION:ClinicalTrials.gov identifier: NCT04925466.

Background:Obesity is a well-established risk factor for obstructive sleep apnea (OSA). We assessed the reciprocal prevalence of obesity and OSA and how it varies by age and sex. Methods:Following a systematic review through March 27, 2025, the final sample included four community-based cohort studies in the US and Switzerland. OSA severity was quantified using the apnea-hypopnea index (AHI, all apneas plus hypopneas with ≥4% oxygen desaturation/hour). Random effects individual participant data (IPD) meta-analyses estimated prevalences. Logistic regression compared odds of OSA across weight groups. Findings:Among 12,860 adults (mean ± SD age: 66.6 ± 7.3 years), 7222 (56.2%) had OSA (AHI ≥5 events/h) and 3309 (25.7%) had obesity (BMI ≥30 kg/m2). IPD meta-analysis showed 31.5% [95% CI: 16.8-48.5] of individuals with OSA had obesity and 44.4% [36.5-52.5] had overweight status (25 ≤ BMI < 30). Among subgroups of individuals with obesity and overweight, 74.3% [63.8-83.5] and 59.8% [46.5-75.7] had any OSA, respectively. Obesity was higher in females than males with OSA, and in younger (<65 years) vs. older individuals. Odds ratios for OSA in subgroups of individuals with overweight and obesity compared to BMI <25 kg/m2 were 2.18 [1.73-2.76] and 4.84 [3.09-6.00], respectively. Interpretation:Our analyses show that most adults with OSA do not have obesity, with 44.4% having overweight and 23.5% having normal weight or underweight. Obesity was more prevalent among females compared to males and in younger individuals (<65 years) compared to older individuals with OSA. Recognizing OSA is not exclusive to obesity highlights the need for personalized treatment plans. Funding:American Academy of Sleep Medicine, National Heart, Lung, and Blood Institute, and Apnimed.

Rationale: Hypoglossal nerve stimulation (HGNS) has emerged as a popular alternative to continuous positive airway pressure (CPAP) for treating obstructive sleep apnea (OSA), yet approximately 30% of patients experience suboptimal responses. High body mass index (BMI) and unfavorable pharyngeal collapse patterns, such as complete concentric palate (CCC) or oropharyngeal lateral wall (OLW) collapse, are associated with lower HGNS efficacy. Understanding the interaction between BMI and pharyngeal collapse site could help optimize patient selection for HGNS. This study tests the hypothesis that the decline in HGNS efficacy with increased BMI is greater with laterally-directed collapse patterns (e.g., complete OLW or complete lateral or concentric palate collapse, CCC), compared to anteroposterior-directed collapse patterns. Methods: We tested this hypothesis using pooled data from two independent HGNS cohorts: a multi-center retrospective cohort (n=338) and a prospective observational cohort from our center (n=377). Both cohorts provided demographics, site of collapse from drug-induced sleep endoscopy (DISE), and baseline and treatment (efficacy polysomnogram) OSA severity. We compared two patterns of pharyngeal collapse: laterally-directed collapse (CCC and complete OLW collapse); and AP-directed (complete collapse of the tongue base, velum [AP-directed only], and epiglottis). To account for multi-site collapse, those with complete tongue collapse were categorized as AP-directed, while those with complete OLW collapse without complete tongue collapse were categorized as laterally-directed, regardless of palate or epiglottic collapse pattern. Those with exclusively partial collapse were grouped separately. Multiple linear regression examined the association between BMI and HGNS efficacy within laterally-directed collapse versus AP-directed collapse (BMI × laterally-directed collapse interaction, primary hypothesis test), adjusting for baseline AHI, partial collapse, and surgical center. Results: Baseline characteristics for analyzed participants were: median [IQR] Age: 60[53,68] years, 487 (68%) men, BMI: 29.1[26.6,31.3]kg/m2, AHI: 32.0[22.6,43.6] events/h, 72 (10%) with laterally-directed collapse, 406 (57%) with AP collapse, and 238 (33%) with partial collapse. HGNS efficacy was weaker in laterally-directed collapse (75.5[50.0,90.1]%) compared to AP (85.7[66.7,95.6]%). The interaction between BMI and laterally-directed collapse was significant (β[95%CI]: −21.8[−47.7, −2.3] per 2SD increase in BMI, p=0.026) and demonstrated that laterally-directed collapse is associated with a greater reduction in efficacy (−29.0[−52.1,−5.9]% per 2SD increase in BMI) than in AP collapse (−7.3[−11.8, −3.0]% per 2SD increase in BMI). Conclusion: Higher BMI is associated with reduced HGNS efficacy in those with laterally-directed collapse patterns, compared to those with AP collapse. These findings highlight the importance of considering pattern of collapse in the context of BMI when selecting candidates for HGNS therapy.

Rationale: Randomized trials of continuous positive airway pressure (CPAP) treatment for obstructive sleep apnea (OSA) have not detected clear evidence of reduced risk of major adverse cardiovascular and cerebrovascular events (MACE) in unselected patients, contrasting with observational studies that have shown cardiovascular benefits of CPAP. Here, we evaluated whether cardiovascular benefit of CPAP occurs preferentially in “high-risk” patients, identified by high levels of OSA-related heart rate acceleration or hypoxemia (effect modification). Methods: In a post-hoc analysis of pooled RICCADSA, ISAACC, and SAVE randomized trials, outcomes were stratified by high-risk OSA status, defined by heart rate response following OSA respiratory events >9.4 beats/min (highest tertile) or oxygen desaturation area-under-baseline (hypoxic burden) >87.1 %.min/hr (highest tertile). Cox mixed models quantified the CPAP treatment effect on MACE (composite including cardiovascular mortality, myocardial infarction, stroke) within high-risk OSA and the difference versus low-risk status (treatment×subgroup interaction, primary hypothesis test). Secondary analyses examined participants without excessive sleepiness (Epworth <11 points) or without increased blood pressure (systolic/diastolic <140/90 mmHg). Other definitions of high-risk were evaluated in sensitivity analyses. Results: In the 3549 participants analyzed, 16.6% and 16.3% met the MACE endpoint with CPAP and usual care respectively. The CPAP treatment effect was greater in participants with versus without high-risk OSA (interaction hazard ratio [iHR]=0.69, 95% confidence interval [CI]: 0.50-0.95, Pinteraction=0.024; overall CPAP benefit within high-risk OSA: hazard ratio [HRhigh-risk] = 0.83[0.66-1.05], Nhigh-risk=1832; CPAP effect within low-risk OSA: HRlow-risk = 1.22 [0.96-1.54], Nlow-risk=1717). Stronger effect modifications were observed after excluding those with excessive sleepiness (iHR=0.59[0.41-0.84]; HRhigh-risk=0.76[0.60-0.98], Nhigh-risk=1509), or those with increased blood pressure (iHR=0.54[0.36-0.81]; HRhigh-risk=0.72[0.54-0.96], Nhigh-risk=1244). Notably, CPAP benefit in high risk OSA was accompanied by an increased hazard in CPAP versus usual care within low-risk OSA (without excessive sleepiness: HRlow-risk = 1.30 [1.01-1.66]; without increased blood pressure, HRlow-risk = 1.33 [1.00-1.76]; Figure 1). Other definitions of high-risk OSA revealed similar findings. Conclusions: High-risk OSA, per OSA-related heart rate acceleration or hypoxemia, modifies the effect of CPAP on MACE, especially in individuals without excessive sleepiness or increased blood pressure. Our study provides novel evidence that a definable subgroup of patients with OSA may benefit from CPAP for major cardiovascular event prevention.

Central sleep apnea is commonly encountered in sleep surgical practices. In this How I Do It, we describe 3 different scenarios in which the off-label use of acetazolamide therapy has been effectively used to treat central sleep apnea in a sleep surgical practice.

Rationale: Patient selection for hypoglossal nerve stimulation (HGNS) for obstructive sleep apnea (OSA) requires assessment of the pharyngeal site of collapse using drug-induced sleep endoscopy (DISE). The current study addresses two key knowledge gaps: First, we prospectively assessed whether, among HGNS candidates, reduced HGNS efficacy is associated with oropharyngeal lateral wall (OLW) collapse (Aim 1). Second, given DISE is a resource-intensive procedure that delays treatment, we evaluate whether a recently-developed non-invasive method for identifying OLW collapse using airflow shapes is associated with reduced HGNS efficacy (Aim 2). Methods: Patients who underwent DISE, HGNS implantation, and follow-up sleep testing were included in Aim 1 (n=347). For Aim 2, airflow data estimating OLW collapse probability (Pr. OLW) were collected during DISE via pneumotachograph (n=133, DISE Flow cohort). In a separate cohort, airflow data was obtained from home sleep testing (Nox A1) via nasal cannula for validation (n=41, HST cohort). Pr. OLW was computed from our validated flow shape analysis model. Linear regression models quantified the association between HGNS efficacy (%AHI reduction from baseline) and DISE-determined OLW collapse (Aim 1) or flow-shape-determined OLW collapse (likelihood score per 2SD; Aim 2), adjusting for baseline AHI. A probability cutoff for OLW collapse was developed to categorize patients into likely HGNS successes (who may skip DISE), and those needing further assessment (e.g., DISE). This cutoff maximized the difference in HGNS efficacy from the DISE Flow cohort and then was validated in the HST cohort. In both cohorts, we quantified the degree to which HGNS efficacy was greater in “likely successes” (Pr. OLW>0.6) versus patients “needing further assessment” using the same modelling approach as above. Results: Compared to non-OLW collapse, patients with DISE-determined OLW collapse had lower HGNS efficacy [95%CI] by -13.9%[-26.7,-3.1]. Increased flow-shape-determined OLW collapse probability was associated with reduced HGNS efficacy in both DISE Flow (-19.4%[-33.1,-7.7]) and HST (-30.3 [-61.6, -7.6]%) cohorts. Further, patients in the needs further assessment group (Pr. OLW>0.60, n=30/133) exhibited lower HGNS efficacy than patients in the likely success group (n=103/133), by -20.6 [-35.7,-7.9]% (Figure 1B). Validation in HST cohort also demonstrated lower HGNS efficacy in the needs further assessment group (n=25/41) versus likely successes (Figure 1C), by -21.7 [-46.3,-3.2], p=0.020. Conclusion: This study prospectively validates OLW collapse as a key factor in HGNS failure and shows that airflow-based identification of OLW collapse can effectively estimate HGNS efficacy, presenting a significant advancement in opportunities for patient selection for HGNS.

BACKGROUND:Patient selection for hypoglossal nerve stimulation (HGNS) for obstructive sleep apnoea (OSA) requires assessment of pharyngeal site of collapse using drug-induced sleep endoscopy (DISE). The current study aims to address two key knowledge gaps. First, we prospectively confirm that, among HGNS candidates, reduced HGNS efficacy is associated with oropharyngeal lateral wall (OLW) collapse (Aim 1). Second, given DISE is a resource-intensive procedure and delays treatment, we evaluate whether a recently developed non-invasive method for identifying OLW collapse using airflow shapes is associated with reduced HGNS efficacy (Aim 2). METHODS:Patients who underwent DISE, HGNS implantation and follow-up sleep testing were included in Aim 1 (n=369) as part of an observational cohort study. For Aim 2, airflow data estimating OLW collapse probability were collected during DISE via a pneumotachograph (n=138; DISE Flow cohort) and from a home sleep test (HST) via nasal cannula for validation (n=46; HST cohort). Linear regression quantified associations between HGNS efficacy (percent reduction in apnoea-hypopnoea index (AHI)) and DISE-determined OLW collapse (Aim 1) or flow shape-determined OLW collapse (probability score per 2sd) (Aim 2), adjusting for baseline AHI. RESULTS:Compared to non-OLW collapse, DISE-determined OLW collapse reduced HGNS efficacy by -18.0% (95% CI -31.9- -6.2%). Increased flow shape-determined OLW collapse probability (Δ2sd) was associated with reduced HGNS efficacy in both DISE Flow (-24.8%, 95% CI -40.4- -11.7%) and HST (-22.7%, 95% CI -50.0- -2.6%) cohorts. CONCLUSION:This study prospectively validates OLW collapse as a key factor in HGNS failure and shows that airflow-based identification of OLW collapse can effectively estimate HGNS efficacy, representing a significant advancement in patient selection for HGNS.

BACKGROUND:OSA pharmacologic interventions like the noradrenergic muscle stimulant atomoxetine have wake-promoting properties. Pimavanserin, a promising serotonin 2A receptor antagonist, may help to counteract atomoxetine's noradrenergic effects by increasing arousal threshold and possibly reducing OSA severity. RESEARCH QUESTION:What is the effect of the combination of pimavanserin and atomoxetine on apnea-hypopnea index (AHI; primary outcome), arousal index, and nadir oxygen saturation (Spo2; secondary outcomes)? STUDY DESIGN AND METHODS:After baseline polysomnography, 18 participants with OSA (AHI > 15 events/h) took pimavanserin plus atomoxetine (34/80 mg; 34/40 mg for the first 3 days) or placebo for 1 week according to a randomized, crossover, 2-period, double-masked clinical trial. Follow-up polysomnography was performed to provide study outcomes. Safety outcomes, subjective sleep quality, and flow-estimated endotypes (using oronasal pneumotachograph flow) also were explored. RESULTS:Eleven and 7 participants were randomized to atomoxetine plus pimavanserin and placebo first, respectively. The combination reduced AHI by 42% (95% CI, 18%-60%) vs placebo, meeting the primary outcome (P < .001). Absolute AHI reduction was 16.9 events/h (95% CI, 8.1-23.6 events/h) more than placebo. Nadir Spo2 and arousal index also were improved, by 5.0% (95% CI, 1%-8%) and 10.9 events/h (95% CI, 2.4-18.1 events/h) vs placebo. Overnight heart rate was increased (+4.8 beats/min; 95% CI, 1.5-8.1 beats/min), but no other change in subjective sleep quality or next-morning vital signs was evident. No increased risk for side effects was observed for the combination vs placebo. Treatment vs placebo improved pharyngeal collapsibility (+7.9% of stable breathing during sleep; 95% CI, 1.6%-14.1% of stable breathing during sleep), reduced loop gain by 20% (0.15; 95% CI, -0.23 to -0.07), and did not reduce the arousal threshold. INTERPRETATION:Our results indicate that pimavanserin with atomoxetine is a strong pharmacologic therapy candidate for OSA. CLINICAL TRIAL REGISTRATION:ClinicalTrials.gov; No.: NCT05350215; URL: www. CLINICALTRIALS:gov.

Background: A recent multi-centric clinical trial, MARIPOSA, examined efficacy, safety and tolerability of AD109, a combination of atomoxetine and aroxybutynin vs. placebo and atomoxetine alone for treatment of sleep apnea. After a 4-week administration, AD109 substantially reduced the apnea hypopnea index (AHI4, 4% desaturation definition for hypopneas) vs. placebo and Patient Reported Outcome Questionnaires (PROMIS)‒fatigue vs. placebo and atomoxetine alone. However, there was variability in AHI reduction (i.e. 44% had an AHI reduction of more than 50%). Here, we sought to identify potential contributors to treatment response with AD109 75-2.5mg. Methods: Response to treatment was defined as a ≥50% reduction in AHI4 from baseline. In secondary analyses, patients were considered responders if the PROMIS‒fatigue also improved. We considered that a low body mass index (BMI) would predict response to treatment (logistic regression). PROMIS‒sleep impairment and ‒fatigue scores were considered as secondary predictors. Analyses included adjustment for baseline AHI4, age and sex. Results: Treatment response was associated with lower baseline BMI (OR [95%CI]: 2.94[1.18‒10.00] per 1SD; N=34), higher baseline PROMIS‒sleep impairment (3.00[1.12‒10.38] per 1SD) and higher PROMIS‒fatigue (2.82[0.96‒12.30]). When PROMIS‒fatigue was added to the response definition, associations were slightly attenuated (BMI: 2.50[1.06-7.69]; PROMIS‒sleep impairment: 1.94[0.76‒5.92; PROMIS‒fatigue: (1.55[0.61‒5.11]). Conclusions: Responses to AD109 appear greatest in patients with reduced obesity and greater sleep apnea symptoms, who may therefore be most suitable for future trials.

Rationale: A low respiratory arousal threshold is a key endotype responsible for obstructive sleep apnea (OSA) pathogenesis. Pimavanserin is an antiserotoninergic capable of suppressing CO2-mediated arousals without affecting the respiratory motor response in animal models, and thus it holds potential for increasing the arousal threshold in OSA and subsequently reducing OSA severity. Objectives: We measured the effect of pimavanserin on arousal threshold (primary outcome), OSA severity, arousal index, and other OSA endotypes (secondary outcomes). Methods: A total of 18 OSA participants were studied in a randomized, double-blind, crossover study. Patients received a single dose of placebo or pimavanserin 34 mg 4 hours before in-lab polysomnography. Airflow was measured with an oronasal mask attached to a pneumotachograph, and ventilatory drive was recorded with an intraesophageal electromyography catheter. Results are presented as mean or median changes (Δ) and 95% confidence intervals (CIs). Results: Pimavanserin did not increase the arousal threshold, nor did it decrease OSA severity or arousal index. It, however, prolonged total sleep time (Δ[confidence interval (CI)], 39.5 [95%CI, -1.2 to 80.1] min). In an exploratory analysis, a subgroup of seven patients who had a 10% or more increase in arousal threshold on pimavanserin exhibited a decrease in AHI4 (hypopneas associated with 4% desaturation) (Δ[CI], 5.6 [95%CI, 3.6-11.1] events/h) and hypoxic burden (Δ[CI], 22.3 [95%CI, 6.6-32.3] %min/h). Conclusions: A single dose of pimavanserin did not have a significant effect on arousal threshold or OSA severity. However, in a post hoc analysis, a subset of patients who exhibited an increase in arousal threshold on pimavanserin showed a small decrease in OSA severity. Thus, if the arousal threshold could be increased with pimavanserin, perhaps with longer dosing to reach higher drug blood concentrations, then the desired effect on OSA severity might be achievable. Clinical trial registered with ClinicalTrials.gov (NCT04538755).