Enhanced Drug-Induced Sleep Endoscopy: Distinguishing Central from Obstructive Apnea

INTRODUCTION

Drug-induced sleep endoscopy (DISE) is a diagnostic procedure frequently employed by sleep apnea surgeons to visualize upper airway collapse in patients with obstructive sleep apnea (OSA). Since its introduction in 1991, findings on DISE have been used to guide surgical OSA management based on the site(s), severity and pattern(s) of upper airway obstruction. Central sleep-disordered breathing events, however, can also be associated with upper airway obstruction. Distinguishing obstructive from central events with purely endoscopic assessment during DISE can be challenging.

CASE REPORT

A 58-year-old non-obese male with snoring, nocturia and excessive daytime sleepiness was diagnosed with moderate obstructive sleep apnea (OSA) (apnea-hypopnea index [AHI] 17 events/hour) on an outside flow-based home sleep study. This claustrophobic patient was referred to consider therapeutic alternatives to continuous positive airway pressure (CPAP) including hypoglossal nerve stimulation (HGNS). His past medical history was notable for rhinitis, cardiomyopathy, atrial fibrillation and congestive heart failure with preserved ejection fraction (55%), all treated medically. Drug-induced sleep endoscopy was performed using the Drug-Induced Sleep Endoscopy with Characterization of Airway Dynamics (DISE-CAD) protocol (UPenn IRB, IRB#833511), which acquires real-time respiratory airflow and airway pressure signals alongside endoscopic image capture, enabling simultaneous anatomical and physiological assessment.¹ Continuous positive airway pressure was also titrated after baseline observation during DISE-CAD to quantify upper airway collapsibility.²

Our patient displayed anteroposterior collapse at the velum, tongue base and epiglottis during baseline breathing at atmospheric nasal pressure under propofol (at a targeted bispectral index range of 50–70). This collapse pattern would have rendered him suitable for HGNS. However, a waxing and waning Hunter-Cheyne-Stokes breathing (HCSB) pattern emerged during CPAP titration from 5 to 9 cmH2O (Figure 1, upper panel) with marked pharyngeal collapse during the hyperpneic rather than apneic phase. This finding prompted a follow-up polysomnogram that showed severe predominantly central sleep apnea with a HCSB pattern (CSA-HCSB) (AHI 32 events/hour; central apnea index 20 events/hour). A prolonged lung-to-finger circulation time of 30–35 seconds, often observed with the central diathesis, was also noted (Figure 2).³ Given the revised diagnosis of CSA, he underwent cardiopulmonary evaluation and proceeded with phrenic nerve stimulation instead of HGNS.

Figure 1:

Pressure-flow signals and videoendoscopic images during Drug-Induced Sleep Endoscopy with Characterization of Airway Dynamics (DISE-CAD).

Marked pharyngeal collapse is seen during the hyperpneic breath (A) as opposed to during the apneic phase (B) when continuous positive airway pressure is applied. Flow and pressure signals parallel one another throughout two disordered breathing cycles, demonstrating continuity and no obstruction across the entire length of the pharynx. CPAP = continuous positive airway pressure, E = epiglottis, Flow = airflow measured at nose with pneumotach, P = soft palate, P_DS = pressure downstream (at retroepiglottic region), P_US = pressure upstream (at retropalatal region), PPW = posterior pharyngeal wall, T = tongue base.

Figure 2:

Four-minute respiratory montage from polysomnogram during non-rapid eye movement sleep.

Characteristic Hunter-Cheyne-Stokes breathing illustrated with central apneas punctuated by waxing and waning hyperpneic breaths and oxyhemoglobin desaturations. CanSnore = snore sensor, PFlow = nasal pressure flow transducer, TFlow = oral thermistor flow, THO = thoracic effort belt, ABD = abdominal effort belt, SpO2 = oxygen saturation.

DISCUSSION

This case illustrates the importance of correlating physiologic and anatomic assessment during DISE. In CSA-HCSB, pharyngeal collapse can occur during the apneic event and/or the hyperpneic phase.⁴ Collapse can occur at the ventilatory nadir in HCSB due to simultaneous reductions in neural drive to upper airway dilators and respiratory pump muscles, masking the central nature of the event. A marked increase in drive during the hyperpneic phase of HCSB can also result in collapse and snoring,⁵ especially after CPAP splinted the airway open during the apneic phase in our patient (Figure 1, lower panel). The presence of pharyngeal collapse on videoendoscopy alone, as during the standard DISE, can unfortunately reinforce a mistaken impression that obstructive sleep-disordered breathing patterns predominate.

As home sleep studies become increasingly common in diagnosing OSA, the physician must remain highly attuned to the possibility of CSA masquerading as OSA, which can be easily misclassified on home sleep studies. Atrial fibrillation and heart failure are common cardiac conditions that predispose to CSA-HCSB and should alert physicians, particularly if CSA-HCSB is suspected on history or DISE. Other conditions associated with CSA-HCSB include stroke and raised intracranial pressure. As sleep surgeons, we understand that diagnostic clarity before surgery is paramount. In our case, a high-quality attended polysomnogram helped classify the respiratory events correctly with major implications for the patient’s ultimate therapy.

The DISE-CAD setup involves a CPAP machine with nasal mask connected to a pneumotachometer to measure decrements in airflow, and airway pressure sensors to measure respiratory effort. Alternatively, respiratory inductance plethysmography belts commonly incorporated in home sleep studies can provide surrogate measures for respiratory effort. These signals can reliably and objectively demonstrate the absence of respiratory effort, which can be a potentially critical assessment in patients undergoing sleep surgery evaluation. In the setting of a standard DISE procedure with endoscopy alone, the sleep surgeon must be sensitive to the absence of respiratory effort suggestive of a central diathesis. Endoscopic clues include a lack of “tugging” on a collapsed pharynx and a lack of effort-related mucus migration, or a motionless thoracoabdominal region.

CONCLUSION

Central sleep apnea can be easily mistaken for OSA as upper airway collapse can appear in both disorders. The sleep surgeon must be certain an obstructive diathesis is present prior to offering surgical intervention. Drug-induced sleep endoscopy affords a unique opportunity to observe simulated sleep-related breathing patterns, and can be further enhanced with additional measurements to clearly distinguish between these two types of apnea.

ABBREVIATIONS

AHI

Apnea-hypopnea index

CPAP

Continuous positive airway pressure

CSA

Central sleep apnea

CSA-HCSB

Central sleep apnea with Hunter-Cheyne-Stokes breathing

DISE

Drug-induced sleep endoscopy

DISE-CAD

Drug-Induced Sleep Endoscopy with Characterization of Airway Dynamics

HCSB

Hunter-Cheyne-Stokes breathing

HGNS

Hypoglossal nerve stimulation

OSA

Obstructive sleep apnea