[Characterastics and determinants of upper airway collapse in OSA patients with different genioglossus negative pressure reflex patterns during DISE].

Objective:To analyze individual differences and influencing factors of upper airway collapse in patients with moderate to severe obstructive sleep apnea（OSA） with different genioglossus negative pressure reflex patterns. Methods:This prospective cohort study started from January 2021 to September 2023, recruited male patients with moderate to severe OSA who were treated at Tsinghua Chang Gung Hospital in Beijing. After completing upper airway CT, each subject underwent drug-induced sleep endoscopy（DISE） with synchronous genioglossus electromyography（GGEMG） and epiglottis level intracavitary pressure monitoring（Pepi）. During spontaneous obstructive apnea events, the peak phasic GGEMG and corresponding inspiratory and epiglottis negative pressure（Pepi） were recorded as pairing data for linear regression analysis to classify GG response patterns. Differences in CT measurements, GGEMG parameters during wakefulness and sleep, and the dynamic changes of upper airway collapse under DISE were compared between the two groups. Statistical analysis was conducted using SPSS 21.0. Results:A total of 35 patients with moderate to severe OSA were analyzed, including 16 patients in the peak GGEMG-Pepi linear group（Group 1） and 19 patients in the peak GGEMG-Pepi non-linear group（Group 2）. There were no significant statistical differences in age, BMI, PSG data, and upper airway CT measurements between two groups. Patients in group 2 showed significant genioglossus（GG） activity defects, which showed a higher complete collapse rate of tongue base airway（Group 1: 16.7% vs Group 2: 30.8%）. Further constructing the regression equation（P<0.001, r=0.974, r²=0.948） for predicting AHI in Group 2 patients showed: AHI=0.229×minimum cross-sectional area of velopharynx airway+2.862×velopharynx airway length+2.456×glossopharynx airway length-0.623×ratio of peak phasic GGEMG to Pepi during awake（R1） -114.897. Conclusion:The peak GGEMG-Pepi classification can effectively distinguish different genioglossus negative pressure reflex patterns for OSA patients. Among them, patients with peak GGEMG-Pepi non-linear mode are more likely to have genioglossus dysfunction, for whom the surgical treatment should be carefully considered.