Airflow Shape Is Associated With the Pharyngeal Structure Causing OSA

Abstract

Background: OSA results from the collapse of different pharyngeal structures (soft palate, tongue, lateral walls, and epiglottis). The structure involved in collapse has been shown to impact non-CPAP OSA treatment. Different inspiratory airflow shapes are also observed among patients with OSA. We hypothesized that inspiratory flow shape reflects the underlying pharyngeal structure involved in airway collapse.

Methods: Subjects with OSA were studied with a pediatric endoscope and simultaneous nasal flow and pharyngeal pressure recordings during natural sleep. The mechanism causing collapse was classified as tongue-related, isolated palatal, lateral walls, or epiglottis. Flow shape was classified according to the degree of negative effort dependence (NED), defined as the percent reduction in inspiratory flow from peak to plateau.

Results: Thirty-one subjects with OSA (mean apnea-hypopnea index score ± SD, 54 ± 27 events/h) who were 50 ± 9 years of age were studied. NED was associated with the structure causing collapse (P < .001). Tongue-related obstruction (n = 13) was associated with a small amount of NED (median, 19; interquartile range [IQR], 14%-25%). Moderate NED was found among subjects with isolated palatal collapse (median, 45; IQR, 39%-52%; n = 8) and lateral wall collapse (median, 50; IQR, 44%-64%; n = 8). The epiglottis was associated with severe NED (median, 89; IQR, 78%-91%) and abrupt discontinuities in inspiratory flow (n = 9).

Conclusions: Inspiratory flow shape is influenced by the pharyngeal structure causing collapse. Flow shape analysis may be used as a noninvasive tool to help determine the pharyngeal structure causing collapse.

Keywords: OSA; flow shape; site of pharyngeal collapse.

Figure 1

A-C, Schematic representation of a posteriorly located tongue. The tongue pushes against the soft palate (A and B) and epiglottis (C), producing bulging of the palate (A and B). The anterior velopharyngeal wall is convex down and results in anteroposterior narrowing, suggesting that the palate is being pushed posteriorly by the tongue (B, velopharyngeal endoscopic view). The tongue base is prominent and obliterates the vallecula (C, oropharyngeal view).

Figure 2

A-C, Thirteen different representative examples of inspiratory flow patterns with corresponding end-expiratory endoscopic views. Single inspiratory flows from each subject are overlaid. The darker tracing represents the ensemble average inspiratory flow. Note that inspiratory flow is more scooped (larger negative effort dependence) among subjects with isolated palatal and lateral wall collapse compared with subjects with posteriorly located tongue. A, Posteriorly located tongue. B, Isolated palate. C, Lateral walls.

Figure 3

A-E, Representative examples of flow shape, Epi, and endoscopic images at end-expiration and peak negative inspiratory pressure. A, Tongue-related obstruction (posteriorly located tongue) associated with stable flow limitation (small negative effort dependence [NED]), with little phasic movement from end-expiration to peak negative pressure can be seen at the velopharynx (A) and oropharynx (B) (examples taken from the same subject). C, D, Isolated palate (C) and lateral wall (D) involvement showing moderate NED associated with more movement of the palate and lateral walls. E, Epiglottic collapse showing large NED and abrupt discontinuities in flow shape associated with lateral folding of the epiglottis. Note that pharyngeal pressure does not decrease as it does in panels A through D. Examples shown in D and E were taken from the same subject at different moments, highlighting the involvement of two different structures causing collapse in the same subject. Epi = pharyngeal pressure.

Figure 4

Representative examples of epiglottic collapse taken from different subjects. Epiglottic collapse is characterized by severe negative effort dependence with and without abrupt discontinuities. Note that during unobstructed breaths, the epiglottis is not wide open and therefore may be more vulnerable to sudden obstruction.

Figure 5

Negative effort dependence according to the pharyngeal structure causing collapse. *P < .05; **P < .01; ***P < .001. Hollow dots represent seven subjects that have the epiglottis and an additional structure involved in collapse. Solid dots represent negative effort dependence of each subject according to the structure involved in collapse. Horizontal lines represent the medians of each group.

Figure 6

Flowchart to identify the pharyngeal structure causing collapse. Small NED (< 34.4%) was able to detect tongue-related obstruction with a sensitivity of 100% and specificity of 92%. Patients with moderate to severe NED (> 34.4%) could have collapse at the lateral walls, palate, or epiglottis. An NED of ≥ 60.5% was able to detect patients with epiglottic collapse with a sensitivity of 89% and specificity of 90%. In addition, the presence of discontinuities among patients with severe NED was able to identify patients with epiglottic collapse because all had abrupt discontinuities in flow. See Figure 3 legend for expansion of abbreviation.