Schroeder JW, Bhandarkar ND, Holinger LD. Synchronous Airway Lesions and Outcomes in Infants With Severe Laryngomalacia Requiring Supraglottoplasty. Arch Otolaryngol Head Neck Surg. 2009;135(7):647-651. doi:10.1001/archotol.125.12.1305
To quantify the prevalence and the impact of synchronous airway lesions identified by endoscopy in infants undergoing supraglottoplasty for severe laryngomalacia (LM).
Tertiary care pediatric hospital.
Sixty patients who underwent supraglottoplasty for severe LM from 2002 to 2006. Patients who underwent preoperative tracheotomy, had previous airway surgery, or did not have 6 months of follow-up were excluded. Fifty-two patients met inclusion criteria.
Supraglottoplasty (with carbon dioxide laser).
Main Outcome Measures
Presence of synchronous airway lesions and their contribution to upper airway obstruction (UAO) and their effect on the postoperative course after supraglottoplasty.
Fifty-eight percent of patients had synchronous airway lesions (SALs), of whom 77% had subglottic stenosis (SGS) and 47% had tracheomalacia, bronchomalacia, or both. Sixty-three percent of all patients required postoperative nonsurgical airway support. Eight patients had residual UAO requiring additional surgical intervention, with 3 revision supraglottoplasties and 7 tracheotomies performed. Infants with neurological conditions had a high rate of surgical intervention (55%; P = .001). Patients with SGS exceeding 35% but without any neurological condition had a prolonged hospital stay (>3.6 days; P = .02) and an 83% incidence (P = .04) of postoperative UAO requiring intubation. Infants with LM with laryngeal edema (LE) alone had increased frequency of postoperative nonsurgical airway support (P = .02) and a prolonged hospital stay of 1 day (P = .01) compared with infants without edema.
There is a high incidence of SALs in patients undergoing supraglottoplasty. Neurological conditions, hypoplastic mandible, SGS greater than 35%, and preexisting LE independently adversely affected the postoperative course.
Laryngomalacia (LM) is the most common cause of pediatric stridor.1 It is characterized by dynamic inspiratory collapse of supraglottic structures associated with shortened aryepiglottic folds, redundant cuneiform and arytenoid mucosa, and an exaggerated omega-shaped epiglottis. Alterations in sensorimotor integrative function and tone of the larynx may be instrumental in pathogenesis.2 Gastroesophageal reflux disease (GERD) and neurological conditions are known to exacerbate airway compromise.3- 5 Treatment is based on the severity of the upper airway obstruction (UAO). Children with severe laryngomalacia causing symptoms such as respiratory distress, cyanosis, or failure to thrive may require surgical intervention at an estimated incidence of approximately 10%.6
Supraglottoplasty is currently the surgical intervention of choice to treat severe LM causing persistent UAO. It is performed with a variety of techniques including carbon dioxide laser, microdebrider, and cold knife instrumentation. Treatment is directed at the specific structures involved in supraglottic collapse and may include lysis of aryepiglottic folds, removal of excess cuneiform and arytenoid mucosa, or epiglottoplasty. Failure of supraglottoplasty to significantly improve airway obstruction is rare but may result in the need for revision supraglottoplasty or tracheotomy.7
The incidence rate of synchronous airway lesions (SALs) in all children with LM has been studied and is estimated to be 19% to 27%.8- 10 The definition of SAL in published studies varies slightly. The frequency of SALs in children undergoing supraglottoplasty has been studied, but specific information on the effect of SALs on postoperative course is lacking.11 Flexible endoscopy is easily and routinely performed in the outpatient setting in children with UAO. This procedure is often sufficient to confirm the diagnosis of LM. When children undergo supraglottoplasty for severe LM, direct laryngoscopy and bronchoscopy are performed as part of a complete airway evaluation. The effectiveness of this examination and the relevance of its findings are evaluated in this study. Specifically, we quantify the prevalence of synchronous airway lesions identified by endoscopy and report their effect on outcomes in children undergoing supraglottoplasty for severe LM.
This is a retrospective study performed at a tertiary care pediatric hospital with approval of the institutional review board. Patients with severe LM who underwent supraglottoplasty at Children's Memorial Hospital in Chicago, Illinois, from 2002 to 2006 were evaluated.
Medical charts of patients who underwent supraglottoplasty owing to severe UAO and/or failure to thrive caused by LM were reviewed. Only patients who underwent carbon dioxide laser–assisted supraglottoplasty were included in this study. Sixty patients met these criteria. Eight patients who had previous airway surgery, preexisting tracheostomy, supraglottoplasty by any other technique, or inadequate documentation of hospital course were excluded, leaving 52 patients for evaluation.
Demographic information including birth history and age at the time of surgery was recorded. All patients were evaluated preoperatively with fiberoptic laryngoscopy to confirm the diagnosis of LM. The presence of laryngeal edema (LE) was noted. Airway fluoroscopy, preoperative swallow study, and dual pH probe studies were not routinely performed. Direct laryngoscopy and bronchoscopy were typically performed at the time of supraglottoplasty. We defined SALs as airway anomalies (other than LM) that may contribute to UAO, including vocal fold motion impairment, subglottic stenosis (SGS), tracheomalacia, and bronchomalacia. The degree of SGS was objectively measured by endotracheal tube size resulting in an air leak of 10 to 25 cm H2O cross-referenced with patient's age on a standard table to determine the percentage of SGS. The actual percentage of SGS rather than Myer-O’Connor-Cotton grade was recorded.12 The approximate degree of UAO from tracheomalacia and bronchomalacia was determined by visual comparison of inspiratory and expiratory lumen diameter. In addition, systemic and congenital conditions contributing to UAO were considered for analysis. These included neurological conditions (cerebral palsy, hypotonia), hypoplastic mandible, and choanal stenosis.
Supraglottoplasty was performed by suspension microlaryngoscopy using a Parson laryngoscope with insufflation anesthetic technique. Incisions were made using a carbon dioxide laser at a setting of 4-W superpulse sequence adapted to a microscope manipulator as previously described.6 Treatment was directed at bilateral cuneiform cartilages, posterior supraglottic mucosa, and the aryepiglottic folds. The children in the study did not undergo epiglottoplasty.
Duration of hospital stay was recorded. All patients were observed at least overnight in the pediatric intensive care unit. The use of any form of nonsurgical airway support was reviewed. These interventions included supplemental oxygen, racepinephrine hydrochloride, heliox (a helium-oxygen mixture), noninvasive positive pressure ventilation (continuous positive airway pressure [CPAP] or bilevel positive airway pressure [BiPAP]), nasal trumpet, and endotracheal intubation. The duration of supplemental oxygen requirement and intubation were recorded.
Differences between groups of patients on the continuous measures (age, duration of supplemental oxygen) were evaluated using a 2-sided t test. For frequencies, significant differences in proportion between groups were tested using a χ2 test. P ≤ .05 was considered significant.
Fifty-two patients met inclusion criteria. Their mean age, obstructive anomalies, and neurological conditions were as follows:
Sixty-three patients required postoperative nonsurgical airway support:
Eight patients required surgical intervention to stabilize the airway. Revision supraglottoplasty was performed in 3 of these patients, although 2 patients with persistent UAO after revision eventually required tracheotomy. Therefore, a total of 7 patients required tracheotomy. Of the patients requiring further surgical intervention, 6 had a neurological condition and 2 had a hypoplastic mandible. The single patient whose UAO resolved with revision supraglottoplasty had a neurological condition. There were 5 patients with a neurological condition who did not require tracheotomy.
The 11 patients with neurological conditions were compared with the other 41 patients without neurological conditions (Table 1). There was no difference between the 2 groups with respect to prevalence of SGS or LE. However, there was a statistically significant association between neurological conditions and both prematurity (P = .02) and tracheomalacia (P = .001). Six of the 11 patients required additional surgical intervention after supraglottoplasty owing to persistent UAO. Patients with a neurological condition also required increased postoperative nonsurgical airway support, including a 41% higher rate of supplemental oxygen requirement (P = .02), a 33% higher rate of intubation (P = .05), and longer hospital stay (>7.3 days; P = .001) compared with patients without such conditions.
Fourteen patients had SGS in the absence of a neurological condition or hypoplastic mandible. Eight patients had SGS of 35% or less, and 6 had SGS greater than 35%. No patient in these 2 groups required additional surgical intervention after supraglottoplasty. However, compared with patients with SGS of 35% or less, patients with SGS greater than 35% had a 50% higher rate of requiring postoperative nonsurgical airway support (P = .05). This included a 75% higher rate of supplemental oxygen (P = .01), a 58% higher rate of intubation (P = .04), a longer period of intubation (>1.5 days), and a longer hospital stay (>3.6 days; P = .02) (Table 2).
Of the 20 patients with LM in the absence of SALs, neurological conditions, and hypoplastic mandible, 12 had mild-to-moderate posterior supraglottic edema and 8 had no edema (ie, LM alone). The patients without LE had uncomplicated postoperative courses with no requirement for airway support and were discharged after an overnight stay. Seven patients with LE (58%) required postoperative nonsurgical airway support (P = .02). In this group of patients, 4 (33%) required supplemental oxygen and 3 (25%) required intubation for a mean duration of 1.3 days. These patients also required, on average, 1 additional day of hospitalization, which was statistically significant (P = .005) (Table 3).
Patients born prematurely underwent supraglottoplasty at a mean age of 8 months, which matched the mean age of all patients evaluated. Of the 13 patients born prematurely, 4 required additional postoperative surgical intervention. All 4 had a neurological condition or hypoplastic mandible. Excluding these patients, there were no clinically significant differences between premature and term infants. Therefore, prematurity was not considered an independent factor that negatively affected postoperative outcomes.
Laryngomalacia is the most common cause of strider in infants. This is generally a self-limited, nonsevere condition that does not require surgical intervention. In most cases, history and physical examination (including flexible fiberoptic laryngoscopy and high-kilovolt anteroposterior and lateral radiographs of the cervical airway) are all that is needed to diagnose LM.9 Routine use of rigid bronchoscopy in this group is not necessary, nor is it recommended. However, in the small group of patients with severe LM who require surgical intervention, complete airway evaluation, including direct laryngoscopy and rigid bronchoscopy, is required.
Studies have reported the incidence of SAL in all patients with LM to be 19% to 27%.8- 10 Infants with severe LM requiring supraglottoplasty have a higher incidence of SALs. Toynton et al11 reported a 47% incidence rate of SALs specifically in patients undergoing supraglottoplasty. In this study, we report a similar incidence rate, and we attribute this to the nature of the patients seen at a tertiary care center. It is possible these lesions contributed to the degree of UAO seen. However, with the exception of 8 patients requiring additional surgical intervention owing to persistent UAO, supraglottoplasty was effective in relieving symptoms.
Subglottic stenosis was a specific SAL contributing to poorer outcome. Our data suggest that SGS greater than 35% is a significant independent risk factor for requiring additional postoperative airway support (P = .01). Mancuso et al9 published several criteria that necessitate rigid endoscopy in the setting of LM. We support the recommendation that all infants with severe LM undergoing supraglottoplasty should be evaluated with rigid endoscopy. During this procedure, objective measurement of SGS should be performed to assist in intraoperative and postoperative planning and treatment.
Posterior supraglottic LE independently increased the need for postoperative airway support in patients with isolated LM. Laryngeal edema was assumed to be caused by laryngopharyngeal reflux. However, 24-hour double-probe pH monitoring was not routinely performed to confirm this hypothesis.13 It has been previously shown that GERD can cause LE and subsequently worsen the clinical presentation of LM.4
Infants with a neurological condition that may contribute to airway hypotonia were more likely to require postoperative airway support and had an increased hospital stay. Previous studies11,14 have similarly demonstrated a negative impact of neurological conditions and associated congenital anomalies on supraglottoplasty outcomes. Our results demonstrated a statistically significant prevalence (P = .001) of tracheomalacia in infants with neurological conditions that may be related to the generalized hypotonic state of these patients. Although 55% of patients with a neurological condition required tracheotomy in this review, the remainder avoided tracheotomy with appropriate postoperative care, including 1 successful revision supraglottoplasty. Therefore, we do not recommend prophylactic tracheotomy in all patients with a neurological condition and severe LM.
The presence of a neurological condition should be differentiated from a subclinical neurological dysfunction that still requires further study as far as its role in the pathogenesis of LM is concerned. Thompson2 has shown that alterations in the sensorimotor integrative function and tone of the larynx are instrumental in pathogenesis. It is not currently known if this process occurs in the same fashion in clinically significant neurological conditions, such as anoxic brain injury with resultant cerebral palsy and, in many cases, airway hypotonia.
The most clinically significant complication of supraglottoplasty that could cause persistent airway obstruction independent of other factors is SGS. Complications had been reported as occurring at a rate of 3% in a previous study but did not occur in our patients. Several recommendations have been previously published to minimize risk of this complication.14
The incidence of SAL in the subset of infants with severe LM requiring supraglottoplasty is clearly higher than when compared with all infants with LM. Specific SAL affects the postoperative course in patients who undergo supraglottoplasty. Subglottic stenosis greater than 35% and pre-existing LE independently increase the need for postoperative airway support and can prolong the postoperative hospital stay. Comorbidities, such as neurological conditions causing hypotonia, will also adversely affect the postoperative course. This may be due to the persistent UAO caused by the higher incidence of tracheomalacia seen in this group. Premature birth was not considered an independent factor that negatively affected postoperative outcomes.
The high incidence of SALs necessitates complete airway evaluation in all patients with severe LM undergoing supraglottoplasty. An increased need for postoperative airway support and a prolonged hospital stay following supraglottoplasty should be anticipated in infants with neurological conditions, hypoplastic mandible, SGS greater than 35%, and preexisting LE because these factors independently adversely affected postoperative course.
Correspondence: James W. Schroeder Jr, MD, Division of Pediatric Otolaryngology, Children's Memorial Hospital, 2300 Children's Plaza, Box 25, Chicago, IL 60614 (email@example.com).
Submitted for Publication: October 20, 2008; accepted November 30, 2008.
Author Contributions: All of the authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Schroeder, Bhandarkar, and Holinger. Acquisition of data: Bhandarkar. Analysis and interpretation of data: Schroeder, Bhandarkar, and Holinger. Drafting of the manuscript: Bhandarkar. Critical revision of the manuscript for important intellectual content: Schroeder, Bhandarkar, and Holinger. Study supervision: Schroeder and Holinger.
Financial Disclosure: None reported.
Previous Presentation: This study was presented at the Combined Otolaryngology Annual Spring Meeting of the Society of Pediatric Otolaryngology; May 4, 2008; Orlando, Florida.