Real-time prediction of disordered breathing events in people with obstructive sleep apnea.

PURPOSE: Conventional therapies for obstructive sleep apnea (OSA) are effective but suffer from poor patient adherence and may not fully alleviate major OSA-associated cardiovascular risk factors or improve certain aspects of quality of life. Predicting the onset of disordered breathing events in OSA patients may lead to improved strategies for treating OSA and inform our understanding of underlying disease mechanisms. In this work, we describe a deployable system capable of performing real-time predictions of sleep disordered breathing events in patients diagnosed with OSA, providing a novel approach for gaining insight into OSA pathophysiology, discovering population subgroups, and improving therapies. METHODS: LArge Memory STorage and Retrieval artificial neural networks with 864 different configurations were applied to polysomnogram records from 64 patients. Wavelet transforms, measures of entropy, and other statistics were applied to six physiological signals to provide network inputs. Approximate statistical tests were used to determine the best performing network for each patient. The most important predictors of disordered breathing events in OSA patients were determined by analyzing internal network parameters. RESULTS: The average optimized individual prediction sensitivity and specificity were 0.81 and 0.77, respectively. Predictions were better than random guessing for all OSA patients. Analysis of internal network parameters revealed a high degree of heterogeneity among disordered breathing event predictors and may reveal patient subgroups. CONCLUSIONS: We report the first practical system to predict individual disordered breathing events in a heterogeneous group of patients diagnosed with OSA. The pattern of disordered breathing predictors suggests variable underlying pathophysiological mechanisms and highlights the need for an individualized approach to OSA diagnosis, therapy, and management.

Automated prediction of apnea and hypopnea, using a LAMSTAR artificial neural network.

RATIONALE: The prediction of individual episodes of apnea and hypopnea in people with obstructive sleep apnea syndrome has not been thoroughly investigated. Accurate prediction of these events could improve clinical management of this prevalent disease. OBJECTIVES: To evaluate the performance of a system developed to predict episodes of obstructive apnea and hypopnea in individuals with obstructive sleep apnea; to determine the most important signals for making accurate and reliable predictions. METHODS: We employed LArge Memory STorage And Retrieval (LAMSTAR) artificial neural networks to predict apnea and hypopnea. Wavelet transform-based preprocessing was applied to six physiological signals obtained from a set of polysomnography studies and used to train and test the networks. MEASUREMENTS AND MAIN RESULTS: We tested prediction performance during non-REM and REM sleep as a function of data segment duration and prediction lead time. Measurements included average sensitivities, specificities, positive predictive values, and negative predictive values. Prediction performed best during non-REM sleep, using 30-second segments to predict events up to 30 seconds into the future. Most events were correctly predicted up to 60 seconds in the future. Apnea prediction achieved a sensitivity and specificity up to 80.6 +/- 5.6 and 72.8 +/- 6.6%, respectively. Hypopnea prediction achieved a sensitivity and specificity up to 74.4 +/- 5.9 and 68.8 +/- 7.0%., respectively. CONCLUSIONS: We report, to our knowledge, the first system to predict individual episodes of apnea and hypopnea. The most important signal for apnea prediction was submental electromyography. The most important signals for hypopnea prediction were submental electromyography and heart rate variability. This prediction system may facilitate improved therapies for obstructive sleep apnea.

Drug-Induced Sleep Endoscopy and Hypoglossal Nerve Stimulation Outcomes: A Multicenter Cohort Study.

OBJECTIVES/HYPOTHESIS: To determine the association between findings of blinded reviews of preoperative drug-induced sleep endoscopy (DISE) and outcomes of hypoglossal nerve stimulation (HNS) for obstructive sleep apnea (OSA). STUDY DESIGN: Cohort study. METHODS: A retrospective, multicenter cohort study of 343 adults who underwent treatment of OSA with HNS from 10 academic medical centers was performed. Preoperative DISE videos were scored by four blinded reviewers using the VOTE Classification and evaluation of a possible primary structure contributing to airway obstruction. Consensus DISE findings were examined for an association with surgical outcomes based on therapy titration polysomnogram (tPSG). Treatment response was defined by a decrease of ≥50% in the apnea-hypopnea index (AHI) to <15 events/hour. RESULTS: Study participants (76% male, 60.4 ± 11.0 years old) had a body mass index of 29.2 ± 3.6 kg/m2 . AHI decreased (35.6 ± 15.2 to 11.0 ± 14.1 events/hour; P < .001) on the tPSG, with a 72.6% response rate. Complete palate obstruction (vs. none) was associated with the greatest difference in AHI improvement (-26.8 ± 14.9 vs. -19.2 ± 12.8, P = .02). Complete (vs. partial/none) tongue-related obstruction was associated with increased odds of treatment response (78% vs. 68%, P = .043). Complete (vs. partial/none) oropharyngeal lateral wall-related obstruction was associated with lower odds of surgical response (58% vs. 74%, P = .042). CONCLUSIONS: The DISE finding of primary tongue contribution to airway obstruction was associated with better outcomes, whereas the opposite was true for the oropharyngeal lateral walls. This study suggests that the role for DISE in counseling candidates for HNS extends beyond solely for excluding complete concentric collapse related to the velum. LEVEL OF EVIDENCE: 3 Laryngoscope, 131:1676-1682, 2021.

Assessment and Management of Postoperative Pain Associated with Sleep Apnea Surgery.

A literature review was conducted regarding the assessment and treatment of postoperative pain following surgery for obstructive sleep apnea (OSA). Given the risks of opioid use by patients with OSA, special attention to opioid risk reduction and avoidance is warranted in this population. The results of this review demonstrate the existence of a body of evidence that supports the use of nonopioid analgesics and nonpharmacologic approaches pain management. Strategies for managing postoperative pain should emphasize the use of local anesthetic infiltration, nonsteroidal antiinflammatory drugs, acetaminophen, topical analgesics, surgical wound cooling, and when necessary, safer opioid medications, such as tramadol and intranasal butorphanol.

Drug-Induced Sleep Endoscopy and Surgical Outcomes: A Multicenter Cohort Study.

OBJECTIVE: To evaluate the association between findings of blinded reviews of preoperative drug-induced sleep endoscopy (DISE) examinations using the VOTE Classification and obstructive sleep apnea (OSA) surgical outcomes in a large multicenter, international cohort. METHODS: Retrospective, multi-center cohort study of adults without tonsillar hypertrophy who underwent pharyngeal surgery for OSA. The study included only participants without enlarged tonsils. Four independent reviewers performed blinded review of preoperative DISE videos using the VOTE Classification system and scoring of a primary structure contributing to airway obstruction. DISE findings were examined for an association with surgical outcomes with univariate analyses and multiple regression. RESULTS: Two hundred seventy-five study participants were included from 14 centers. Mean age was 51.4 ± 11.8 years, and body mass index was 30.1 ± 5.2 kg/m2 . There was moderate interrater reliability (kappa = 0.40-0.60) for DISE findings. Oropharyngeal lateral wall-related obstruction was associated with poorer surgical outcomes (adjusted odds ratio (AOR) 0.51; 95% CI 0.27, 0.93). Complete tongue-related obstruction was associated with a lower odds of surgical response in moderate to severe OSA (AOR 0.52; 95% CI 0.28, 0.98), with findings that were similar but not statistically significant in other analyses. Surgical outcomes were not clearly associated with the degree and configuration of velum-related obstruction or the degree of epiglottis-related obstruction. Surgical response was associated with tonsil size and body mass index (inversely). CONCLUSION: DISE findings concerning the oropharyngeal lateral walls and tongue may be the most important findings of this evaluation technique. LEVEL OF EVIDENCE: 2B Laryngoscope, 129:761-770, 2019.

Establishing a danger zone: An anatomic study of the lingual artery in base of tongue surgery.

OBJECTIVES/HYPOTHESIS: To contrast the changes in measurement of the hypoglossal/lingual artery neurovascular bundle (HLNVB) to constant surface landmarks in the base of tongue (BOT) during surgically simulated retraction versus resting anatomic position, and to identify a safe zone for BOT robotic surgery to avoid injury to the HLNVB. STUDY DESIGN: Human cadaver study. METHODS: Five fresh-frozen head and neck complexes were obtained, and seven HLNVBs were dissected. A microcaliper was used to measure the distance from the HLNVB to constant surface landmarks in resting and surgically simulated positions using a Feyh-Kastenbauer retractor. RESULTS: Measurements from foramen cecum to palatoglossus muscle (P < 0.042) was significantly different when comparing anatomical to surgically simulated positions. Importantly, the location of the lingual artery in reference to the surface landmarks measured was dramatically altered with tongue retraction. With retraction, the branches of the dorsal lingual artery were not encountered posterior to a horizontal line between midway circumvallate papilla (mCVP). CONCLUSION: Measurements of the HLNVB to surface landmarks in the BOT differs significantly between resting and a surgically simulated tongue position. The dorsal branch of the lingual artery seems more superficial in the BOT than previously described. A safe zone may exist posterior to an imaginary horizontal line between mCVP; however, further studies are needed to confirm this. LEVEL OF EVIDENCE: NA Laryngoscope, 127:110-115, 2017.

Brown Norway and Zucker Lean rats demonstrate circadian variation in ventilation and sleep apnea.

STUDY OBJECTIVES: Circadian rhythms influence many biological systems, but there is limited information about circadian and diurnal variation in sleep related breathing disorder. We examined circadian and diurnal patterns in sleep apnea and ventilatory patterns in two rat strains, one with high sleep apnea propensity (Brown Norway [BN]) and the other with low sleep apnea propensity (Zucker Lean [ZL]). DESIGN/SETTING: Chronically instrumented rats were randomized to breathe room air (control) or 100% oxygen (hyperoxia), and we performed 20-h polysomnography beginning at Zeitgeber time 4 (ZT 4; ZT 0 = lights on, ZT12 = lights off). We examined the effect of strain and inspired gas (twoway analysis of variance) and analyzed circadian and diurnal variability. MEASUREMENTS AND RESULTS: Strain and inspired gas-dependent differences in apnea index (AI; apneas/h) were particularly prominent during the light phase. AI in BN rats (control, 16.9 ± 0.9; hyperoxia, 34.0 ± 5.8) was greater than in ZL rats (control, 8.5 ± 1.0; hyperoxia, 15.4 ± 1.1, [strain effect, P < 0.001; gas effect, P = 0.001]). Hyperoxia reduced respiratory frequency in both strains, and all respiratory pattern variables demonstrated circadian variability. BN rats exposed to hyperoxia demonstrated the largest circadian fluctuation in AI (amplitude = 17.9 ± 3.7 apneas/h [strain effect, P = 0.01; gas effect, P < 0.001; interaction, P = 0.02]; acrophase = 13.9 ± 0.7 h; r (2) = 0.8 ± 1.4). CONCLUSIONS: Inherited, environmental, and circadian factors all are important elements of underlying sleep related breathing disorder. Our method to examine sleep related breathing disorder phenotypes in rats may have implications for understanding vulnerability for sleep related breathing disorder in humans.