Obstructive sleep apnea is an increasingly recognized problem. In cases of obstructive sleep apnea in which there is a complicated medical history, predisposition to airway obstruction at multiple levels, or persistent sleep apnea despite an apnea-treating surgical procedure, dynamic sleep fluoroscopy is a useful adjunct to endoscopic evaluation of the airway. This article describes the indications and technique for dynamic sleep fluoroscopy and the anatomic sites and common causes of airway obstruction.
Dr. Donnelly is an Associate Professor and Dr. Strife is a
Professor in the Department of Radiology, and Dr. Myer is a
Professor in the Division of Otolaryngology at the Children's
Hospital Medical Center, Cincinnati, OH.
It is estimated that up to 3% of all children, approximately 2
million in the United States alone, are affected by obstructive
sleep apnea syndrome.
The most common cause of obstructive sleep apnea is enlarged
adenoid and palatine tonsils in otherwise healthy children. Other
causes of obstructive sleep apnea include craniofacial anomalies,
congenital syndromes (particularly Down's syndrome and
achondroplasia), mucopolysaccharidosis, and prior surgery on the
Many of these patients are predisposed to airway obstruction at
Polysonography is helpful in differentiating between central versus
obstructive causes of sleep apnea.
However, it provides no accurate information concerning the
anatomic level of obstruction in patients with obstructive sleep
apnea. In cases of obstructive sleep apnea in which there is a
complicated medical history or persistent sleep apnea following a
surgical procedure performed to treat sleep apnea, dynamic sleep
fluoroscopy has been shown to be a useful adjunct to endoscopic
evaluation, affecting management decisions in more than 50% of
It is particularly helpful in identifying dynamic abnormalities of
the airway, such as functional collapse, as compared with static,
fixed obstructions. Despite this, dynamic sleep fluoroscopy in
children has received little attention in the imaging literature.
In 1979, Felman et al
described their cinefluoroscopic technique in 9 children using
sleep deprivation. We have performed more than 80 dynamic sleep
fluoroscopy procedures in children using sedation. The purpose of
this review is to describe the indications and technique for
dynamic sleep fluoroscopy and the anatomic sites and common causes
for airway obstruction.
When children present with symptoms of airway obstruction,
imaging evaluation typically includes frontal and lateral
radiographs of the airway and chest and flexible fiberoptic
laryngoscopy. If extrinsic tracheal compression is suspected,
cross-sectional imaging with computed tomography (CT) or magnetic
resonance imaging (MRI) is usually performed. If an intrinsic or
dynamic problem of the trachea is suspected, direct laryngoscopy
and bronchoscopy under general anesthesia is usually performed.
For obstructive sleep apnea, one of the advantages of dynamic
sleep fluoroscopy over flexible fiberoptic laryngoscopy is the
ability to evaluate the entire airway simultaneously when the child
is sleeping. We use sleep fluoroscopy to evaluate those children
with complex medical histories who are at increased risk of
multilevel airway obstruction.
Multiple disease processes are associated with an increased risk
of multilevel airway obstruction. In Down's syndrome, airway
obstruction can occur secondary to macroglossia, lymphoid
hyperplasia, congenitally narrow nasopharynx, laryngomalacia,
congenital subglottic stenosis, tracheobronchomalacia, or tracheal
Clinical evidence of airway obstruction may be secondary to any, or
all, of these potential sites. Children with neuromuscular
disorders are also at risk for airway collapse at multiple levels
secondary to muscular hypotonia.
Other children at risk for multilevel airway obstruction include
those with congenital craniofacial anomalies, such as Pierre Robin
syndrome, or metabolic disorders, such as the
In patients who have had previous surgery and have persistent sleep
apnea, the problem may be related to residual obstruction secondary
to the original cause or secondary to sequelae of the surgical
manipulation of the airway.
In some patients who have had tracheotomies, there may be
difficulty with decannulation. These children may have episodes of
apnea or respiratory distress during sleep secondary to development
of granulation tissue or localized tracheomalacia at the surgical
site or recurrence of the primary problem.
At our institution, indications for evaluation with dynamic
sleep fluoroscopy include: 1) persistent symptoms of sleep apnea
despite normal findings on flexible fiberoptic laryngoscopy; 2)
persistent symptoms of sleep apnea after a single site of
obstruction has been identified and appropriately treated; 3)
potential for obstruction at more than one site within the upper
airway because of either previous surgery or an underlying
abnormality; and 4) difficulty decannulating a patient following
The study is performed and monitored by a pediatric radiologist.
Patients are prepared for the procedure according to our
departmental sedation program guidelines.
During the procedure, a radiologist, radiology technologist, and
radiology nurse are present. Food and drink are withheld for 4 to 8
hours prior to the examination to decrease the risk of aspiration.
Patients are sedated with either oral chloral hydrate (70 to 100
mg/kg) or intravenous pentobarbital (3 mg/kg, with repeat dosing up
to a total of 7 mg/kg) depending upon patient age. During the
entire procedure and sedation recovery, respiratory rate, heart
rate and rhythm, and blood oxygen saturation are monitored using
transcutaneous pulse oximetry. The child's parents are allowed and
encouraged to attend the procedure to reduce the child's anxiety
and to verify if the sleep patterns observed are typical of those
that occur at home.
The studies are performed in a room equipped with lateral
fluoroscopy. The children are imaged in supine position with
lateral fluoroscopy. The fluoroscopic portions of the examinations
are videotaped with simultaneous audiotaping so that fluoroscopic
findings can be correlated with episodes of oxygen desaturation or
noisy breathing. Physical observations, which are noted and
correlated with fluoroscopic findings, include respiratory effort,
thoracic wall motions, and episodes of apnea. Fluoroscopic
evaluation is performed for approximately 10 to 20 seconds at areas
of anatomic interest when signs of airway occlusion occur. The
fluoroscopic evaluation of a child with sleep apnea should be
performed at three specific sites: the level of the base of the
tongue (oropharynx), the hypopharynx, and the intrathoracic
trachea. Rarely, these three areas may be seen simultaneously. In
larger children, the intrathoracic trachea may need to be evaluated
separately from the hypopharynx. Evaluation of the oropharynx and
hypopharynx are performed with the arms positioned at the child's
sides. Downward pulling on the arms improves visualization of the
neck. Evaluation of the intrathoracic trachea is best performed
with the arms extended above the child's head. To limit radiation
dose, a maximum of 2 minutes of total fluoroscopic time is
Certain maneuvers may be performed during the sleep fluoroscopy
to further evaluate the obstruction. When present, tracheotomy
tubes may be capped in order to see if the patient develops sleep
apnea when the artificial airway is bypassed. The tracheotomy tube
may also be removed to evaluate for underlying tracheomalacia,
which can be masked when the tracheotomy tube is present and
physically prevents the trachea from collapsing. In cases in which
a child has a tracheotomy tube that is to be removed intentionally
or occluded during the study, an otolaryngologist is present to
perform these maneuvers. When an area of airway obstruction is
encountered, the effect of treatments, such as positive pressure
breathing, on decreasing or eliminating the obstruction can be
Normal findings and commonly encountered
One of the advantages of sleep fluoroscopy is the evaluation of
dynamic motion abnormalities in addition to static fixed causes of
obstruction. In the normal sleeping child, there is little or no
motion of the pharynx and trachea. Any dynamic motion of these
structures encountered during sleep should be considered abnormal.
Commonly encountered dynamic abnormalities include glossoptosis,
pharyngeal collapse, laryngomalacia, and tracheomalacia. It must be
stressed that such dynamic abnormalities can occur at multiple
sites or in conjunction with fixed causes of airway
The first line of treatment for many of these dynamic causes of
obstructive sleep apnea is the use of positive-pressure airway
devices during sleep.
Some of these types of obstructive sleep apnea will decrease with
increasing age. Therefore, if positive-pressure ventilation can
relieve the symptoms, it may be the only necessary therapy until
the child outgrows the condition. Knowledge of the specific
abnormality is important because the odds of positive-pressure
therapy being helpful and the odds of the child outgrowing the
condition are different for each specific entity. Therefore, the
length of trial for conservative therapy may be influenced. In
addition, when conservative management fails, the specific types of
surgery that can be performed differ for each of the types of
Glossoptosis is defined as abnormal posterior motion of the
tongue during sleep. It is seen most commonly in children with
neuromuscular abnormalities, because of an abnormal decrease in
It can also be associated with macroglossia and micrognathia. On
fluoroscopy, the tongue "falls" posteriorly during sleep, abutting
the velum (soft palate) and posterior wall of the pharynx,
obstructing the airway (figure 1).
Glossoptosis can be difficult to detect with endoscopic evaluation.
Surgical interventions to either reduce the volume of the tongue or
reposition the mandible have been described for those cases
refractory to medical management.
Pharyngeal collapse is another commonly encountered cause of
obstruction in this population and, like glossoptosis, can be
difficult to detect endoscopically. On fluoroscopy, the anterior
wall of the pharynx moves posteriorly and the posterior wall moves
anteriorly (figure 2). This differs from glossoptosis in which only
the tongue moves posteriorly. With pharyngeal collapse, the
posterior pharyngeal wall, velum, and tongue oppose each other,
causing nasopharyngeal and oropharyngeal obstruction.
Laryngomalacia and tracheomalacia
Laryngomalacia and tracheomalacia are defined as abnormal
collapse of the larynx or trachea during breathing secondary to
lack of normal structural integrity of the underlying cartilage.
With both, the collapse typically occurs during inspiration. Both
the larynx and trachea should be relatively still during sleep, and
any motion should be considered abnormal. With laryngomalacia,
there is inferior indrawing of the pharynx, and the epiglottis
buckles and infolds over the tracheal inlet (figure 3), secondary
to a lack of adequate cartilaginous support. Tracheomalacia can be
focal or diffuse. With tracheomalacia, the trachea cyclically
decreases in caliber (figure 4). Typically, the anterior wall of
the trachea bows and collapses posteriorly more prominently than
the posterior wall moves anteriorly. Tracheomalacia can occur as a
primary weakness of the tracheal cartilage or secondary to
extrinsic compression, such as by anomalous vascular structures or
Enlargement of the adenoid tonsils is one of the more common
components of obstructive sleep apnea in children. However,
determination of what size constitutes an abnormally enlarged
adenoid has been a subject of debate. Several studies have
addressed the range of normal sizes of the adenoid tissues during
The expected size of the adenoid tonsils changes with age. In
newborns, no adenoid tissue may be appreciated at imaging.
There is a rapid proliferation of the adenoid tissues during
infancy with a plateau in size varying from 2 to 14 years of age.
The most typical age of maximal size is 7 to 10 years, at which
time the adenoid tissues may range from 10 to 15 mm in diameter on
a lateral (or sagittal) image.
Beginning in the second decade, the adenoids begin to decrease in
size and continue to do so throughout adulthood.
Dynamic sleep fluoroscopy is a useful adjunct to endoscopy in
the evaluation of dynamic abnormalities of the airway. This study
aids in identifying the actual site of airway obstruction,
particularly when there may be multiple potential causes.