Laryngomalacia: a classification system and surgical treatment strategy.
Laryngomalacia, the most common congenital laryngeal anomaly, is not a single disease entity but rather a variety of entities along a spectrum of underlying pathophysiologies. Based on our study of 10 children who were surgically treated for laryngomalacia in an urban tertiary care center, we have developed a system of classifying laryngomalacia on the basis of its different underlying pathophysiologic processes. Type 1 laryngomalacia is characterized by a foreshortened or tight aryepiglottic fold. Type 2 disease is defined by the presence of redundant soft tissue in the supraglottis. The type 3 designation applies to cases caused by other etiologies, such as underlying neuromuscular disorders. While the three types are not mutually exclusive, each should be considered as a separate disease entity with a final common clinical presentation. Each type requires a specific approach to surgical repair.
Laryngomalacia is the most common cause of neonatal stridor, accounting for one-half to two-thirds of all cases of noisy breathing in this age group. (1) Laryngomalacia usually becomes symptomatic during the first 2 weeks of life, (2) and it usually resolves by the time a child is 12 to 18 months old. (1) Only 10% of cases require surgical intervention, generally to treat apnea or failure to thrive. (1) The underlying defect that causes laryngomalacia remains unknown, but it is believed to involve neuromuscular weakness, cartilaginous inadequacy, or an anatomic abnormality such as redundant arytenoid mucosa.
Various classification schemes for laryngomalacia have been proposed, and assorted treatment modalities and surgical procedures have been described. However, few authors have attempted to mesh the classifications and their treatments to develop a treatment algorithm. In this article, we describe our efforts to tailor the surgical procedure according to the specific type of laryngomalacia.
Patients and methods
We established a hypothesis that the surgical treatment of laryngomalacia specifically directed at the underlying etiology would be effective and sufficient. To test our theory, we developed a classification system for laryngomalacia based on its underlying etiology (figure), and then we determined the most appropriate surgical procedure for each diagnosis.
Type 1 laryngomalacia is characterized by a foreshortened or tight aryepiglottic fold, and we treated it by dividing the folds with scissors. Type 2 disease is defined by the presence of redundant soft tissue in the supraglottis, and it was treated by resecting the redundant mucosa with sharp instruments. The type 3 designation applies to cases caused by other etiologies, such as underlying neuromuscular disorders and posterior epiglottic collapse; it was treated with a tracheostomy.
We studied our hypothesis in 10 patients--8 boys and 2 girls aged 3 months to 3 years (median: 6 mo).
Two of the 10 patients had been born prematurely with bronchopulmonary dysplasia and gastroesophageal reflux (table). The most common reason for surgical intervention was respiratory distress (n = 8), defined as worsening stridor and retraction beyond what is normally seen in laryngomalacia. Six patients experienced failure to thrive, 3 had apnea, and 2 had cyanosis.
Five patients had isolated type 1 laryngomalacia, 3 had isolated type 2 laryngomalacia, 1 patient had a combination of types 1 and 2, and 1 patient had a combination of types 1 and 3. No patient had isolated type 3 laryngomalacia. Other findings noted at the time of direct laryngoscopy and bronchoscopy included posterior laryngeal edema in 4 patients and subglottic stenosis in 2 patients.
Supraglottoplasty was performed on 9 patients; 6 patients underwent an aryepiglottic fold division, and 4 underwent an arytenoid mucosal resection (1 patient underwent both procedures). Tracheostomy was performed on 2 patients. once as a primary procedure and once following the failed resection of bilateral arytenoid redundant mucosa.
The three types of laryngomalacia we describe do not represent a novel classification system. In fact, our classifications reflect the three abnormalities that McSwiney et al in 1977 noted could be present, either separately or in combination. (3) They wrote that (1) the epiglottis could be long and curled upon itself and prolapse posteriorly on inspiration, (2) the aryepiglottic folds could be too short, and (3) the arytenoids could be more bulky than normal and prolapse forward on inspiration. They also noted a high incidence of laryngomalacia among children with cerebral palsy and mental retardation, although they did not specify the particular type of hiryngomalacia in these particular patients. In our experience, patients with neurodevelopmental disorders generally have a posterior prolapsed epiglottis; when surgical intervention is required in such cases, tracheostomy rather than supraglottoplasty is the procedure of choice.
Other authors have advocated different approaches to the surgical treatment of laryngomalacia. In a series of 40 surgical cases of laryngomalacia. Prescott advised surgeons to focus on three supraglouic elements during the physical examination: the epiglottis, the aryepiglottic folds, and the mucosa over the corniculate cartilages on the arytenoids. (4) He found an omega-shaped epiglottis in 27 patients (67.5%), with epiglottal prolapse into the laryngeal inlet in 13 patients (32.5%). All patients had a short aryepiglottic fold and edematous mucosa over the corniculate cartilages that prolapsed anteriorly. Prescott advocated excision of a V wedge from the center of the aryepiglottic folds as the surgical procedure of choice. He reported resecting the mucosa over the corniculate cartilages only ii it was edematous, and in fact he did perform this procedure on every one of the patients he described in that study. He excised the lateral margins of an omega-shaped epiglottis in 32 patients (80.0%) and noted that no patient had persistent epiglottic prolapse requiring epiglottoplasty.
Nussbaum and Maggi defined two types of laryngomahlcia based on criteria important to pulmonologists. (5) The first type is isolated laryngomalacia as defined by McSwiney et al, (3) who noted that this entity is usually detected in early infancy. The second type is laryngomalacia associated with any other bronchoscopic findings, including gastro-esophageal reflux.
In describing a series of 115 patients, Roger et al distinguished three types of laryngomalacia: complete, predominantly posterior, and isolated anterior. (6) The complete type, which was seen in two-thirds of their patients, was characterized by an omega-shaped epiglottis, shortened aryepiglotlic folds, and redundant supraglottic mucosa that may or may not involve the mucosa of the cuneiform cartilages. Most of the other patients had the predominantly posterior type with redundant arytenoid mucosa. Only 2 patients had isolated anterior laryngomalacia with posterior swaying of the epiglottis.
Shah and Wetmore, in designing a reporting form for laryngomalacia, separated the disorder into three types based on the principal site of anatomic collapse. (7) In their series of 10 patients, 4 had posterolateral laryngomalacia with redundant aryepiglottic folds, and 6 had only posterior laryngomalacia with excess arytenoid mucosal or cartilaginous bulk; no patient had anterior laryngomalacia involving the epiglottis. Two of their patients required supraglottoplasty; the laryngomalacia type and surgical technique in these 2 cases were not specified.
Another classification system--in which laryngomalacia is designated as type A, B, or C--is cited in the Taiwanese literature. (8,9) In type A disease, the cuneiform cartilages are redundant and they prolapse during inspiration. In type B laryngomalacia, a long tubular epiglottis curls back upon itself during inspiration. In type C, the flaccid epiglottis prolapses posteriorly against the posterior pharyngeal wall or vocal folds during inspiration.
Nielson et al developed an empiric scale of 0 to 8 for rating laryngomalacia. (10) The arytenoids and the epiglottis are graded separately on a scale of 0 to 4 points, and the grades are added together. The arytenoids may show no collapse (0), subtle collapse (1), collapse that obscures 25 to 50% of the true vocal folds (2), collapse that obscures 75% of the true vocal folds (3), or collapse that obscures 100% of the true vocal folds (4). The epiglottis may show no folding (0), slight lengthwise folding (1), moderate folding without contact of the lateral edges of the epiglottis (2), folding with intermittent contact of the lateral edges (3), or folding with continuous contact with or overlap of the lateral edges (4). This scale provides a potential measure of clinical severity that can complement other systems that classify specific anatomic locations such as the aryepiglottic folds.
In a comprehensive review of laryngomalacia published in 1999, Olney et al classified both surgical and nonsurgical cases into three types. (1) Type 1 laryngomalacia was characterized by prolapse of the mucosa overlying the arytenoid cartilages, type 2 involved foreshortened aryepiglottic folds, and type 3 involved posterior displacement of the epiglottis. Types 1,2, and 3 laryngomalacia were identified in 57, 15, and 13% of their patients, respectively; another 15% had a combination of types, usually types 1 and 2. Our system is similar to theirs, which they established midway through their study about the same time that we created our system. Just as we did, Olney et al used their classification system to direct the type of supraglottoplasty they performed. Type 1 patients underwent excision of the redundant mucosa over the posterolateral arytenoids, type 2 patients underwent division of the aryepiglottic folds, and type 3 patients underwent an epiglottopexy to the base of the tongue. The surgical success rate was 78%. In our series and in the series by Olney et al, (1) all surgical patients who had posterior epiglottic prolapse (type 3) ultimately required a tracheostomy to secure the airway. Also in our series, both of the patients with bronchopulmonary dysplasia required a tracheostomy.
The development of our classification system was motivated by our belief that different anatomic etiologies of laryngomalacia should be treated with different surgical procedures. In many previous reports, authors advocated that all surgical laryngomalacia be treated the same way, regardless of etiology. We believe that such a strategy carries the potential for unnecessary interventions.
Some authors advocate dividing only the aryepiglottic folds, be it with sharp instrumentation or with a laser. (11) In a series of 115 patients, Garabedian et al reported a 98% success rate, with 7 patients requiring additional procedures and 2 failures requiring tracheostomy. (12) On the other hand, some authors advocate resecting only the redundant mucosa. For example, Polonovski et al designed a suction test that involved placing an aspiration cannula into the laryngeal inlet to assess the amount of supraglottic collapse and to ascertain how much and which tissue to resect. (13) While they frequently resected mucosa from the aryepiglottic fold and occasionally from the lateral edge of the epiglottis, they did not divide the aryepiglottic fold, claiming that two-thirds of patients who undergo such division eventually require a revision procedure. Still other authors advocate both division of the aryepiglottic folds and resection of the arytenoid mucosa regardless of the anatomic site responsible for the laryngomalacia. Zalzal et al described using scissors to trim the aryepiglottic folds, lateral edges of the epiglottis, and the mucosa over the arytenoids and corniculate cartilages. (14) Similarly, Marcus et al reported using scissors to trim the obstructing mucosa from the lateral edge of the epiglottis, aryepiglottic folds, and arytenoid cartilage in all patients, even though they saw two distinct patterns of laryngomalacia in their patients; half exhibited anteromedial collapse of aryepiglottic folds and cuneiform cartilage, while the other half manifested only anteromedial collapse of the mucosa overlying the arytenoids. (15)
Jani et al (16) performed a suction test similar to the one described by Polonovski et al (13) to assess how much redundant arytenoid mucosa to resect, but unlike Polonovski et al, they always excised the aryepiglottic fold. Roger et al began by sectioning only the aryepiglottic folds, but they later also resected the redundant arytenoid mucosa, occasionally with the corniculate cartilage; of 115 cases, 4 required an epiglottopexy for significant epiglottic swaying. (6) Finally, Kelly and Gray used the laser to vaporize the cuneiform cartilage and an adjacent wedge of the aryepiglottic fold, but they did so only unilaterally; only 17% of their patients required the same contralateral procedure in 2 to 3 months. (17)
We have taken an approach that is similar to Kelly and Gray's by performing supraglottic redundant mucosa resections unilaterally. In only 1 of our patients did we find that a bilateral resection was required, and even this proved to be insufficient as the patient ultimately required a tracheostomy. It is important to note that this patient was 1 of the 2 patients who had been born prematurely with bronchopulmonary dysplasia; supraglottoplasty was successful in all 8 of our full-term infants.
There are authors who advocate tailoring the supraglottoplasty to the structural etiology of the laryngomalacia without relying on a classification system. Zeitouni and Manoukian insert a rigid suction into the supraglottis, assess the excess redundant portions, and ablate them with a C[O.sub.2] laser; the treated areas "usually include the aryepiglottic folds, lateral edges of the epiglottis and the corniculate cartilages." (2) Occasionally, they need to resect only the mucosa over the arytenoids. They also excise a wedge out of the aryepiglottic folds if they are too short, and they write that they have never needed to perform an epiglottopexy. Remacle et al always resect the aryepiglottic fold from the arytenoids to the lateral edge of the epiglottis; if necessary, surgery includes the endolaryngeal mucosa of the arytenoids and/or the lateral edge of the epiglottis. (18)
In conclusion, it is possible, of course, to perform selected procedures in accordance with an algorithm that does not involve a classification system. However, we find that a classification system focuses the surgeon's attention preoperatively and intraoperatively to the particular offending anatomy, thereby avoiding the use of a shotgun surgical approach.
(1.) Olney DR, Greinwald JH Jr., Smith RJ, Bauman NM. Laryngomalacia and its treatment. Laryngoscope 1999; 109:1770-5.
(2.) Zeitouni A, Manoukian J. Epiglottoplasty in the treatment of laryngomalacia. J Otolaryngol 1993;22:29-33.
(3.) McSwiney PF, Cavanagh NR Languth P. Outcome in congenital stridor (laryngomalacia). Arch Dis Child 1977;52:215-18.
(4.) Prescott CA. The current status of corrective surgery for laryngomalacia. Am J Otolaryngol 1991; 12:230-5.
(5.) Nussbaum E, Maggi JC. Laryngomalacia in children. Chest 1990;98:942-4.
(6.) Roger G, Denoyelle F, Triglia JM, Garabedian EN. Severe laryngomalacia: Surgical indications and results in 115 patients. Laryngoscope 1995; 105:1111-17.
(7.) Shah UK, Wetmore RF. Laryngomalcia: A proposed classification form. Int J Pediatr Otorhinolaryngol 1998;46:21-6.
(8.) Liu HC, Lee KS, Hsu CH, Hung HY. Neonatal vallecular cyst: Report of eleven cases. Changgeng Yi Xue Za Zhi 1999;22:615-20.
(9.) Lee KS, Chen HL, Yang CC, et al. Surgical management of severe laryngomalacia. J Taiwan Otolaryngol Soc 1997;32:235-40.
(10.) Nielson DW, Ku PL, Egger M. Topical lidocaine exaggerates laryngomalacia during flexible bronchoscopy. Am J Respir Crit Care Med 2000;161 : 147-51.
(11.) Seid AB, Park SM, Kearns MJ, Gugenheim S. Laser division of the aryepiglottic folds for severe laryngomalacia. Int J Pediatr Otorhinolaryngol 1985;10:153-8.
(12.) Garabedian EN, Roger G, Denoyelle F, Triglia JM. Severe laryngomalacia: Surgical indications and results. Pediatr Pulmonol Suppl 1997;16:292.
(13.) Polonovski JM, Contencin R Francois M, et al. Aryepiglottic fold excision for the treatment of severe laryngomalacia. Ann Otol Rhinol Laryngol 1990;99:625-7.
(14.) Zalzal GH, Anon JB, Cotton RT. Epiglottoplasty for the treatment of laryngomalacia. Ann Otol Rhinol Laryngol 1987;96:72-6.
(15.) Marcus CL, Crockett DM, Ward SL. Evaluation of epiglottoplasty as treatment for severe laryngomalacia. J Pediatr 1990; 117:706-10.
(16.) Jani P, Koltai P, Ochi JW, Bailey CM. Surgical treatment of laryngomalacia. J Laryngol Otol 1991; 105:1040-5.
(17.) Kelly SM, Gray SD. Unilateral endoscopic supraglottoplasty for severe laryngomalacia. Arch Otolaryngol Head Neck Surg 1995:121:1351-4.
(18.) Remacle M, Bodart E, Lawson G, et al. Use of the CO2-laser micropoint micromanipulator for the treatment of laryngomalacia. Eur Arch Otorhinolaryngol 1996;253:401-4.
David J. Kay, MD, MPH; Ari J. Goldsmith, MD
From the Center for Pediatric Otolaryngology-Head and Neck Surgery, Boynton Beach, Fla. (Dr. Kay), and the Department of Otolaryngology, SUNY Health Science Center at Brooklyn (N.Y.) (Dr. Goldsmith).
Reprint requests: David J. Kay, MD, Center for Pediatric Otolaryngology Head and Neck Surgery, 10301 Hagen Ranch Rd., Boynton Beach, FL 33437. Phone: (561) 881-5236; fax: (561) 736-5662; e-mail: firstname.lastname@example.org
Originally presented in part at the annual meeting of the American Broncho-Esophagological Association; May 14, 2001; Palm Desert, Calif.
Table. Patient data Birth Medical Surgical Age/sex history history indication 3 mo/M Normal FTT 8 mo/M Normal GER FTT, respiratory distress 3 yr/M Normal Hypotonia Respiratory distress 5 mo/M Normal FTT 4 mo/M Normal Respiratory distress 7 mo/M Normal FTT, respiratory distress, apnea 7 mo/M Normal Respiratory distress 4 mo/F Normal Respiratory distress, apnea 4 mo/M Premature BPD, GER FTT, respiratory distress, cyanosis 3 yr/F Premature MRCP, BPD, FTT, respiratory distress, GER cyanosis, apnea LM Other Age/sex type findings Procedure 3 mo/M 1 Resection of bilateral AE folds 8 mo/M 1 Resection of bilateral AE folds 3 yr/M 1 Resection of bilateral AE folds 5 mo/M 1 Resection of bilateral AE folds 4 mo/M 1 SS Resection of bilateral AE folds 7 mo/M 1, 2 PLE Resection of bilateral AE folds, resection of unilateral redundant mucosa 7 mo/M 2 SS Resection of unilateral redundant mucosa 4 mo/F 2 PLE Resection of unilateral redundant mucosa 4 mo/M 2 PLE Initial: resection of bilateral redundant mucosa; later: tracheostomy 3 yr/F 1, 3 PLE Tracheostomy LM = laryngomalacia; FTT = failure to thrive; AE = aryepiglottic; GER = gastroesophageal reflux; SS = subglottic stenosis; PLE = posterior glottic edema; BPD = bronchopulmonary dysplasia; MRCP = mental retardation and cerebral palsy.