Personalizing obstructive sleep apnea management: associating symptom subtype to objective sleep traits and physiological biomarkers

点击翻译按钮获取中文摘要

ABSTRACT Obstructive sleep apnea (OSA), characterized by repeated partial (hypopnea) and complete (apnea) obstruction of the upper aiways during sleep, is a broad diagnostic category with its complex phathophysiology and variable clinical presentations. Using cluster-analysis techniques, we recently identified three distinct OSA subtypes, based on predominant symptoms: (1) Insomnia (presenting with difficulty sleeping with little daytime sleepiness); (2) Excessively sleepy (presenting with excessive drowsines during the daytime, but few complaints about disturbed sleep); and (3) Minimally Symptomatic. Currently, Positive Airway Pressure (PAP) is the first-line treatment for all OSA patients, regardless of the symptom presentation. PAP is expected to improve symptoms by eliminating respiratory events; however, 50-72% of patients with comorbid insomnia and 18-55% of Sleepy patients still suffer from residual symptoms after PAP treatment, suggesting individual differences in symptom responses to the first-line treatment and a critical need for secondary symptom management strategies. Better understanding of physiological mechanisms underlying the 3 symptom subtypes and differential treatment responses will inform future personalized secondary treatments, including behavioral and pharmacologic interventions. The current proposal will directly evaluate these mechanisms using both an existing clinical sample from multiple sleep centers throughout the world and a newly recruited prospective sample. First, we will investigate the underlying physiological signatures of the symptom subtypes at diagnosis in a large database of OSA patients (n = 853) recruited from the Sleep Apnea Global Interdisciplinary Consortium (SAGIC). Second, we will prospectively recruit and follow new patients (n = 360, n = 120 for each subtype) to validate these physiological signatures at baseline, identify patterns of symptom responses to the first-line PAP treatment, and evaluate the physiological and clinical predictors of symptom response within each subtype. To better characterize physiological and clinical signatures, we will leverage novel electroencephalogram (EEG)-derived variables, including the odds ratio of product (ORP, a continuous index of sleep depth), as well as conventional measures from polysomnography (PSG). For the prospective observational study, diagnostic and follow-up PSG after 3 months of PAP treatment will be conducted in all participants. In addition, 2-week actigraphy and sleep questionnaires will be used to capture at-home sleep characteristics and behaviors. Symptoms will be assessed at diagnosis and monthly over follow-up. We will also obtain and bank fasting blood samples for biomarker and Omics (i.e., genomics, proteomics, and metabolomics) studies to explore molecular and biological mechanisms of OSA symptom subtypes in the future. Overall, this study will help us understand the underlying physiology of OSA symptom subtypes and predictors of better treatment response. Results will inform future symptom management strategies, personalized by patients' physiological and clinical information, for each subgroup.