Drug-Induced Sleep Endoscopy in Children

1

Introduction

Over the past decade there has been an evolution in the care for children with obstructive sleep apnea (OSA). Although adenotonsillectomy (T&A) has been regarded as typically successful in treating pediatric OSA, recent publications raise questions regarding its efficacy. Depending on the outcome measure selected, success can be as low as 27% (Apnea/Hypopnea Index [AHI] <1) or as high as 78% (AHI <5). A randomized prospective study on T&A outcomes (CHAT) also demonstrated an overall success of only 79%. Success was defined as a resultant AHI <2 and an obstructive apnea index of <1 event an hour. Given these less favorable outcomes for T&A, pediatric drug-induced sleep endoscopy (DISE) has become increasingly popular over the past 15 years. The major components of DISE are as follows: patient selection, anesthesia, endoscopy, classification of findings, and interpretation of data. Interpretation of findings is beyond the scope of this chapter.

2

Patient Selection

With the proliferation of pediatric DISE publications, it has become apparent that there is wide variability in when and how it is performed. The majority of DISE procedures within our institutions are performed for children with residual OSA after a T&A or for a child with significant OSA but very small tonsils where the surgeon does not believe that the tonsils/adenoid are solely responsible for the obstruction. Table 73.1 outlines the variety of DISE indications. At this time, there are no outcome studies to facilitate patient selection. Although most people obtain a sleep study before DISE, a careful analysis of the polysomnogram (PSG) findings has not been conducted to determine if there are PSG factors that should be a contraindication for DISE. Because none of the anesthetic agents replicate rapid eye movement (REM) sleep, one needs to be cautious of DISE findings who have only REM obstruction.

3

Anesthesia

There is no consensus on the anesthetic technique. Multiple studies have been published using a variety of agents. The ideal agent will provide adequate analgesia under near physiologic conditions and avoid significant airway collapse, respiratory depression, and cardiovascular complications. It should have a predictable duration of action and a smooth awakening.

Unlike adults, most children will not tolerate an insertion of an intravenous line while awake. For pediatric DISE, there is the additional cofounding variable of the inhalational anesthetic. Nitrous is quickly metabolized, whereas sevoflurane lingers longer. Sevoflurane has dose-dependent effects on genioglossus activity. With increasing depth of sevoflurane anesthesia, there is a decrease in phasic inspiratory genioglossus activity but no difference in tonic activity.

Unless one delays endoscopy until the inhalational agent's effect has dissipated, one may not obtain an accurate representation of the child's true obstructive pattern. Unfortunately, there is no ideal agent that replicates all stages of sleep and has no effect on airway dynamics. Ehsan et al. summarizes the variety of anesthetic agents utilized for DISE. Dexmedetomidine is an alpha 2-adrenoreceptor agonist with minimal respiratory effects that has been shown to mimic a physiologic state similar to stage 2 sleep. Propofol and ketamine are common anesthetic agents for DISE. Unfortunately, propofol has been associated with an alinear increase in upper airway collapsibility. Under certain conditions, a propofol technique may provide spurious information due to a disproportionate decrease in genioglossus activity. Ketamine has the potential of increasing airway tone. Theoretically, dexmedetomidine may be the best agent because it has minimal effect on respiratory control. Of major importance, none of the current agents simulate REM sleep. Future outcome studies will be necessary to determine which anesthetic technique will provide accurate information to facilitate surgical selection.