David L. Mandell and Deepak Mehta
For those clinicians who have dedicated themselves to the field of pediatric upper airway obstruction, subglottic stenosis is a well-known disorder, representing the most common laryngotracheal anomaly requiring tracheostomy in infants.1 The unique knowledge, skills, and procedures required to comprehensively manage this clinical entity have been a significant factor in the emergence of pediatric otolaryngology as a distinct subspecialty unto itself. Despite the broadening experience with airway anomalies, its treatment remains challenging and ever-evolving, necessitating constant development of new techniques and surveillance of changing trends.
The subglottis is the narrowest portion of the pediatric laryngotracheal airway, with a diameter of 4.5 to 5.5 mm in full-term newborns, and of just 3.5 mm in preterm newborns.1 The subglottis is also notoriously unforgiving when instrumented by an endotracheal tube, given that it is the only site within the upper airway which is completely encircled by a ring of cartilage (the cricoid). It is estimated that the majority of cases of pediatric subglottic stenosis are acquired, as a result of injury from an indwelling endotracheal tube. Conversely, it has been estimated that as few as 5% of all cases of subglottic stenosis are congenital.2 Of course, the true prevalence of congenital cases are unknown, since any subglottic stenosis that is found after endotracheal intubation is by definition deemed to be acquired, and it is not possible to know if a preexisting congenital stenosis might have been present.
The mechanism by which an endotracheal tube may precipitate the development of subglottic stenosis is believed to be related to pressure necrosis from the tube. Subglottic mucosal ulceration occurs, followed by eventual cricoid cartilage exposure, chondritis, and necrosis. Granulation tissue appears as a result of this cascade of events and the injured tissue can subsequently mature into a fibrotic scar.3Other factors that might potentiate the development of subglottic stenosis in this setting include excessive tube movement and/or repeated extubations/intubations. Elements that have been traditionally associated with laryngeal inflammation and poor wound healing, such as bacterial infection, laryngopharyngeal reflux, and poor nutritional status, are also likely involved in this pathological process.
Acquired subglottic stenosis became an increasingly encountered entity as a direct consequence of the introduction and subsequent popularity of long-term endotracheal intubation as a viable option for maintaining life support for premature newborns in the 1960s.4 As these patients eventually graduated from the intensive care unit (ICU) and their pulmonary status improved with time, subglottic stenosis was often the only remaining reason for the ongoing need for a tracheostomy tube. The desire to achieve tracheostomy decannulation in this population of patients gave rise to rapid developments in the field of laryngotracheal reconstruction in the 1970s.5
Fearon and Cotton are credited with performing the first reliably successful open pediatric airway reconstructions by expanding the lumen of the cricoid cartilage and placing costal cartilage grafts to maintain airway patency.6,7Results were dramatic, and over the course of time it was found that there were no long-term adverse effects on laryngeal growth.7 However, decannulation is not possible in all patients, and the success of the operation was less robust for the most severe, high-grade stenoses. In 1993, Monnier et al described the cricotracheal resection operation, in which the stenotic segment was resected entirely, followed by an end-to-end tracheal-to-laryngeal anastomosis, with high decannulation rates.8 When examining the outcomes of all of these open surgical reconstructive techniques as a whole, the overall pediatric decannulation rate is estimated to be 91%.5 These reconstructive approaches have withstood the test of time and remain the core surgical procedures used to this day for most cases of mature, acquired subglottic stenosis.
Current improvements in endotracheal tube design and management (including using an appropriately sized tube that allows a leak at 20 cm water pressure),3 have resulted in the incidence of subglottic stenosis falling to less than 1%.9 However, prolonged endotracheal intubation is still considered the primary cause of acquired subglottic stenosis in pediatric patients today.
Clinical Evaluation and Diagnosis
Most patients with acquired subglottic stenosis will present either as infants in the ICU who have failed repeated endotracheal extubation attempts, or as children already with tracheostomy tubes who present in the outpatient setting with families interested in pursuing tracheostomy decannulation. Children with congenital subglottic stenosis, which is generally less severe than acquired stenosis, may present with recurrent croup.
Direct microlaryngoscopy with rigid bronchoscopy is the standard mechanism for assessing subglottic stenosis. The length, severity, consistency (soft vs. hard), and degree of inflammation and granulation tissue (or lack thereof) of the stenosis should be noted, as these findings have therapeutic implications.3 The severity of a mature subglottic stenosis can be “graded” using the Myer-Cotton grading chart, in which the largest endotracheal tube that can be placed with an air leak between 10 and 25 cm water pressure is chosen and compared with the expected normal-sized tube for the patient's age.10 Based on this system, Grade I = 0 to 50% obstruction, Grade II = 51 to 70% obstruction, Grade III = 71 to 99% obstruction, and Grade IV = no detectable lumen (Fig. 16.1).10
With regards to congenital subglottic stenosis, it is believed that the most common endoscopic finding is an elliptical cricoid, in which the transverse diameter is less than the anteroposterior diameter.1Another recognized pattern of congenital subglottic stenosis involves the first tracheal ring being trapped within the cricoid cartilage.1 Other anomalies, such as laryngeal cleft, may coexist with congenital subglottic stenosis and should be ruled out.1
Figure 16.1 Myer-Cotton grading system of laryngeal stenosis.10
Reprinted with permission from: Myer CM III, O'Connor DM, Cotton RT. Proposed grading system for subglottic stenosis based on endotracheal tube sizes. Ann Otol Rhinol Laryngol 1994;103(4 Pt 1):319–323.
During endoscopy, the cricoarytenoid joints should be palpated to test for passive mobility. If the arytenoid cartilages do not move freely along the cricoarytenoid joints, posterior glottic stenosis is likely present, resulting from interarytenoid and/or cricoarytenoid joint scarring. Failure to recognize posterior glottic stenosis preoperatively can lead to failure of laryngotracheal reconstruction.
Dynamic airway assessment is also required to assess vocal fold mobility, laryngomalacia, and tracheomalacia.3 Dynamic assessment can be performed with flexible laryngoscopy with the patient wide awake (including in the office setting and/or rigid bronchoscopy with spontaneous respiration in the operating room (or ICU) setting.11
When assessing a potential candidate for laryngotracheal reconstruction, the effect of the airway stenosis on the child's overall quality of life should be determined. The amount of respiratory compromise and pulmonary reserve of a potential laryngotracheal reconstruction candidate should be assessed. Pediatric pulmonologists are often asked to provide preoperative clearance and approval for a reconstruction procedure, helping to ensure that the patient's lungs are healthy enough to allow for the rigors of the reconstruction and the expected permanent removal of the tracheostomy tube.
Laryngopharyngeal reflux of gastric contents is felt by many to have the ability to compromise laryngotracheal wound healing after surgery. As such, it is considered routine practice to attempt to diagnose and treat such reflux before surgery. Diagnostic techniques can include 24-hour dual-probe pH studies, esophagoscopy with biopsy, contrast esophagram and/or impedance probe testing (which can identify nonacid reflux).11 Eosinophilic esophagitis, if present and untreated, is also felt to be a likely contributing factor to laryngeal inflammation and failure of reconstructive pediatric airway surgery, and thus its possible presence should be investigated with preoperative esophageal biopsy.11,12 Laryngopharyngeal reflux can be treated with H2 blockers and proton-pump inhibitors, and if unresponsive, may require Nissen fundoplication.1
The swallowing status of the patient should be determined preoperatively, and assessment for potential aspiration should be undertaken, either with modified barium swallow study, and/or a functional endoscopic evaluation of swallowing. Aspiration is not necessarily a contraindication to reconstructive surgery, but identifying aspiration may allow for more appropriate postoperative expectations, and feeding therapies geared toward managing and minimizing aspiration can be instituted.
Correctable upper airway obstruction (e.g., adenotonsillectomy for adenotonsillar hypertrophy, maxillofacial surgery for midface or mandibular hypoplasia, or supraglottoplasty for severe laryngomalacia) can be performed before laryngotracheal reconstruction if needed.
Open Surgical Techniques
Mild cases of subglottic stenosis (Grade I) often do not require open surgical reconstruction, as they frequently do not cause significant upper airway obstruction.1 Mature, fibrotic subglottic stenoses that are causing markedly symptomatic airway obstruction and/or are associated with tracheostomy dependence can be well-managed with either augmentation (laryngotracheal reconstruction) or resection (cricotracheal resection) techniques.
Laryngotracheal reconstruction, in which the cricoid cartilage is divided anteriorly and/or posteriorly to allow for cartilage graft augmentation, is generally accepted as the standard of care in most cases of mature pediatric laryngo-tracheal stenosis that require intervention.3 Costal cartilage is the most commonly utilized grafting material, although cartilage grafts from many other sites can be used. Thyroid alar cartilage grafts have risen in popularity recently, primarily due to their ease of harvest.3 However, thyroid alar grafts are limited by the size and thickness of the graft and therefore are useful in only selected cases.
Lower grade stenoses (such as Grade II and mild Grade III) may be managed by dividing the anterior cricoid and placing a single anterior cartilage graft. However, if such stenoses are also associated with posterior glottic stenosis (with impaired mobility of the true vocal folds bilaterally), the cricoid should also be split posteriorly, with a posterior cartilage graft placed in addition to the anterior one.
If an augmentation procedure is performed for a Grade IV or a high Grade III stenosis, combined anterior and posterior cartilage grafting is the preferred technique. However, some authors feel that this more severe category of stenosis is more amenable to cricotracheal resection, in which diseased subglottic tissue that might not support cartilage grafting can be completely resected.11 However, if the stenotic segment is too long, or if it is too close to the inferior surface of the true vocal folds (within 1 to 2 mm), anterior and posterior cartilage graft augmentation may still be the procedure of choice.11 If a high-grade subglottic stenosis coexists with posterior glottic stenosis, a cricotracheal resection with a posterior cricoid split and posterior cartilage graft can be performed.11
Anterior Cricoid Split
Cricoid split (anterior laryngotracheal decompression) can be used in newborns with subglottic stenosis.13 Infants must have an isolated subglottic stenosis that is the cause of repeated failed extubation attempts, and must have adequate cardiopulmonary reserve. Typically, the anterior cricoid split operation is performed in a single stage, with endotracheal intubation for 4 or 5 days afterwards.1 Some authors prefer to use a thyroid ala cartilage graft during cricoid split procedures, turning the procedure into a laryngotracheal reconstruction.1 Recently, endoscopic anterior cricoid split has been described with good success.14
Single-Stage Laryngotracheal Reconstruction
A single-stage procedure is one in which, after the reconstruction, the patient leaves the operating room without a tracheostomy tube. The airway is stented with an endotracheal tube, and when the tube is removed (usually 1 week later), the reconstruction is presumably complete and well-healed.3 Depending on surgeon's preference, microlaryngoscopy and bronchoscopy may be performed just before extubation, and if the reconstruction is felt to be healing well, the endotracheal tube can either be removed at that time, or replaced with a smaller diameter tube which can be removed later in the ICU setting whenever the patient is ready and fully weaned off of the ventilator.
Single-stage procedures can be considered when only an anterior graft is used, with or without a posterior cricoid split.3 However, with better intensive care this has been successfully used for posterior grafts as well. A single-stage reconstruction is less likely to be successful in patients with other synchronous airway lesions, particularly tracheomalacia. Single-stage procedures should also be avoided in patients whose anatomy makes reintubation technically difficult should an emergency arise.
One of the major drawbacks of performing laryngotracheal reconstruction in a single stage is the need for very labor-intensive and demanding postoperative care. Most require significant doses of sedatives and analgesics, which can be associated with complications such as weakened respiratory drive and medication withdrawal. The surgeon should be aware of complications such as atelectasis and pneumonia, tube blockage with secretions, and tube dislodgement. Also, the possibility exists that a new tracheotomy may need to be performed if the patient is not able to remain extubated.
Multistage Laryngotracheal Reconstruction
When laryngotracheal reconstruction is performed in multiple stages, the initial reconstruction is supported by an indwelling stent while the tracheostomy site remains patent and cannulated.15 A T-tube can be used for this purpose; such a tube serves the dual purpose of stenting opening the airway while simultaneously allowing for respiration via the stoma. T-tubes can become occluded if too small in diameter, and thus are generally not used under the age of 4 years. Alternatively, a short stent can be used; this type of stent is considered “short” because it ends just proximal to the tracheostomy stoma. Such a stent is typically sutured into place above the stoma, allowing uninhibited ongoing use of traditional tracheostomy tubes. Whichever type of stent is used, it can remain in place as long as is necessary, can be removed at a later date while maintaining the tracheostomy site intact, and tracheostomy decannulation can then proceed in a relatively leisurely fashion whenever the patient is deemed ready.
Staging the reconstruction is a wise choice in the following situations: (1) Grade IV and high Grade III subglottic stenosis, in which the reconstruction is more complex and likely to take longer to heal; (2) patients with poor respiratory reserve who may not be easy to wean off of the ventilator after prolonged intubation; (3) patients with multiple synchronous airway lesions; and (4) patients with compromised wound healing, in which case long-term stenting is desirable.3
In the cricotracheal resection procedure, the anterior and lateral portions of the cricoid cartilage are excised, whereas the posterior cricoid plate is preserved.8 The trachea inferior to the resection is then anastomosed to the remaining cricothyroid segment. A pedicled tracheal mucosal graft can be used to line the posterior cricoid plate. The risk of dehiscence can be minimized by employing maneuvers to reduce anastomotic tension, such as suprahyoid release, tracheal mobilization, and chin-to-chest sutures. The addition of cricotracheal resection to the armamentarium of procedures for subglottic stenosis has further improved the success rate of decannulation, especially for the highest grade stenoses.5
Routine postoperative care after open airway reconstruction typically involves prophylaxis with broad-spectrum antibiotics, with particular attention to prophylaxis against bacteria that may be associated with failure of airway surgery such as Pseudomonas aeruginosa and Staphylococcus aureus).11 Antireflux medication (typically proton-pump inhibitors) is routinely given as well. Sedation and analgesia are titrated to achieve the goals of maintaining patient comfort and minimizing undesired patient movement, while simultaneously trying to prevent cardiopulmonary depression and decrease the likelihood of sedative withdrawal, which can complicate the process of weaning young patients off mechanical ventilation.16
During the early postoperative period after any open airway reconstruction, most authors recommend routine endoscopic surveillance of the airway under anesthesia, so that wound healing can be assessed and steps can be taken to address potential problems (such as granulation tissue formation or early restenosis that may be amenable to dilatation) before mature restenosis occurs.11
With the rousing success of the laryngotracheal reconstruction procedures introduced in the 1970s, older techniques such as laryngotracheal dilatation with bougienage, which had been attempted sporadically in the first half of the 20th century, fell by the wayside.5 However, endoscopic techniques, and in particular laryngotracheal dilatation procedures, have been making a comeback. This recent trend in pediatric airway management has been spurred on by the modern progression toward minimally invasive surgery, the desire to avoid prolonged postoperative ICU stays and their related complications, and the ever-increasing skill level and advanced instrumentation acquired by pediatric airway surgeons.
Currently, it appears that endoscopic management of subglottic stenosis is a reasonable option in the following scenarios: (1) to prevent endotracheal tube-related subglottic injury from progressing to a mature stenosis when active subglottic inflammation and granulation tissue are identified; (2) to attempt to maintain airway patency and promote appropriate wound healing when a reconstructed subglottis is showing signs of inflammation and restenosis in the early postoperative period; (3) to temporarily improve a relatively mature stenotic subglottic airway (with dilatation), thus potentially allowing more time to pass for an infant to grow and become medically stabilized before an open reconstruction is performed; and (4) to possibly avoid tracheotomy altogether, especially with balloon dilatation in the very young infant, or in cases of a soft subglottic stenosis.
The current thinking is that endoscopic techniques for laryngotracheal stenosis are best suited for less severe and less mature stenoses.17 Although endoscopic posterior cricoid split with costal cartilage grafting has been described,18 most endoscopic techniques do not involve cartilage grafting.
Any technique to dilate the stenotic pediatric subglottis may yield similar initial results, but the theoretical benefit of balloon dilatation over bougienage or rigid bronchoscopic dilatation is that the balloon applies dilating forces radially, thus avoiding mucosal shearing which could promote poor wound healing and restenosis. Balloon dilatation can be attempted with angioplasty balloons and Fogarty catheters, but with the recent surge in interest in this technique, pediatric airway-specific balloons are starting to be manufactured (most notably with a lumen that allows ventilation to continue even while the balloon is being deployed). A 70% success rate has been reported with balloon dilatation for acquired Grade II or Grade III Cotton subglottic stenosis in infants in the ICU.19 After dilation, adjuvant therapies such as topical steroid or mitomycin C application are often employed.
Over the past quarter century, lasers have been used for a variety of purposes in the pediatric airway, including removal of laryngeal and suprastomal granulation tissue and glottic webs. For low-grade (e.g., Grade I or II) subglottic stenosis, the stenotic segment can be lasered in a staged fashion, one quadrant at a time with several weeks of healing in between treatments, or the stenosis can be lysed with radial laser incisions followed by immediate balloon dilatation.17 Traditionally, it has been felt that management of pediatric subglottic stenosis with a laser is most appropriate for noncircumferential stenoses and for lesions that are no longer than 1 cm in length.17
The microdebrider is a thin, hollow metal tube that houses a spinning blade with simultaneous suction, allowing for precise and rapid removal of airway lesions (especially granulation tissue and cysts) with the ability to avoid damage to surrounding normal tissue. This instrument removes blood and debris while collecting tissue for histopathology and avoids the thermal injury and airway fires that can be associated with lasers. Given its larger size, however, it cannot be used through a ventilating bronchoscope, and thus its applications are generally limited to the endolarynx and subglottis. A laryngeal radiofrequency ablation (coblation) wand has also recently been introduced and may have similar applications to the microdebrider in the pediatric airway.
Adjuvant Endoscopic Therapies
The unpredictability of wound healing is a constant thorn in the side of pediatric airway surgeons; poor wound healing can undermine an initially successful procedure for laryngotracheal stenosis. Modulation of airway wound healing in a favorable fashion can be attempted with topical pharmacotherapy.
Mitomycin C is an antibiotic derived from the Streptomyces caespitosus bacteria. This pharmacological agent interferes with postsurgical scar formation at the molecular level.17 It is usually applied topically at a concentration of 0.4 mg/mL for a few minutes after airway surgery has been performed. It appears to be a useful adjunct after endoscopic cold, laser, or balloon procedures, with a good safety profile, although its clinical utility is not universally agreed upon. Topical application of corticosteroids can be used in a similar fashion to minimize edema and granulation tissue formation. Topical or nebulized antibiotics can be used, in combination with systemic antibiotics, to mitigate any potential adverse effect on wound healing caused by bacterial infection.
Summary and Future Directions
The specialty of pediatric airway surgery has now reached a point where nearly all cases of laryngotracheal stenosis can be repaired by various techniques and most patients can achieve decannulation. Looking toward the future, clinicians are increasingly devoting more attention to other outcomes of laryngotracheal surgery besides decannulation, including voice and swallowing function. A normal or near-normal voice is achieved in only about 50% of children after laryngo-tracheal reconstruction20; this is an important component of these patients’ quality of life which has traditionally been overlooked.
The recent rise in popularity of balloon dilatation is part of a larger trend toward performing fewer major open reconstructions with prolonged ICU stays, thus promoting faster recoveries and possible avoidance of postoperative stenting and endotracheal intubation. Endoscopic techniques are now being used more regularly to prevent mature stenoses from developing in infants who are found to have early endotracheal tube damage in the ICU setting. The future holds promise for new techniques, such as robotic pediatric laryngeal surgery, which may have the added benefits of steady, precise movements in small spaces, elimination of line-of-sight limitations by using angled endoscopes and flexible robotic instruments, and allowance of 3D-vision capabilities.21 In place of cartilage grafting for open procedures, research is ongoing to find stable porous biomaterials that might maintain airway structure while enhancing wound healing by promoting well-vascularized mucosal growth within the reconstructed airway lumen.22
A variety of techniques are required, sometimes in the same patient, and constant assessment and reassessment of the stenosis has to be performed.5 The old adage, “decision making is at least as important as the actual surgery,” rings particularly true in the field of pediatric airway surgery.3
1. Schroeder JW Jr, Holinger LD. Congenital laryngeal stenosis. Otolaryngol Clin North Am 2008;41(5):865–875, viii
2. Werkhaven JA, Beste D. Diagnosis and management of pediatric laryngeal stenosis. Otolaryngol Clin North Am 1995;28(4): 797–808
3. Boardman SJ, Albert DM. Single-stage and multistage pediatric laryngotracheal reconstruction. Otolaryngol Clin North Am 2008;41(5):947–958, ix
4. McDonald IH, Stocks JG. Prolonged nasotracheal intubation. A review of its development in a pediatric hospital. Br J Anaesth 1965;37:161–173
5. Santos D, Mitchell R. The history of pediatric airway reconstruction. Laryngoscope 2010;120(4):815–820
6. Fearon B, Cotton R. Surgical correction of subglottic stenosis of the larynx: preliminary report of an experimental surgical technique. Ann Otol Rhinol Laryngol 1972;81:508–513
7. Cotton R. Management of subglottic stenosis in infancy and children: review of a consecutive series of cases managed by surgical reconstruction. Ann Otol 1978;87:649–657
8. Monnier P, Savary M, Chapuis G. Partial cricoid resection and with primary tracheal anastomosis in infants and children. Laryngoscope 1993;103:1273–1283
9. Walner DL, Loewen MS, Kimura RE. Neonatal subglottic stenosis—incidence and trends. Laryngoscope 2001;111(1): 48–51
10. Myer CM III, O'Connor DM, Cotton RT. Proposed grading system for subglottic stenosis based on endotracheal tube sizes. Ann Otol Rhinol Laryngol 1994;103(4 Pt 1):319–323
11. de Alarcon A, Rutter MJ. Revision pediatric laryngotracheal reconstruction. Otolaryngol Clin North Am 2008;41(5):959–980, x
12. Dauer EH, Ponikau JU, Smyrk TC, Murray JA, Thompson DM. Airway manifestations of pediatric eosinophilic esophagitis: a clinical and histopathologic report of an emerging association. Ann Otol Rhinol Laryngol 2006;115(7):507–517
13. Rethi A. An operation for cicatricial stenosis of the larynx. J Laryngol Otol 1956;70(5):283–293
14. Horn DL, Maguire RC, Simons JP, Mehta D. Endoscopic anterior cricoid split with balloon dilation in infants with failed extubation. Laryngoscope 2012;122(1):216–219
15. Wootten CT, Rutter MJ, Dickson JM, Samuels PJ. Anesthetic management of patients with tracheal T-tubes. Paediatr Anaesth 2009;19(4):349–357
16. Hammer GB. Sedation and analgesia in the pediatric intensive care unit following laryngotracheal reconstruction. Paediatr Anaesth 2009;19(Suppl 1):166–179
17. Lando T, April MM, Ward RF. Minimally invasive techniques in laryngotracheal reconstruction. Otolaryngol Clin North Am 2008;41(5):935–946, ix
18. Inglis AF Jr, Perkins JA, Manning SC, Mouzakes J. Endoscopic posterior cricoid split and rib grafting in 10 children. Laryngoscope 2003;113(11):2004–2009
19. Durden F, Sobol SE. Balloon laryngoplasty as a primary treatment for subglottic stenosis. Arch Otolaryngol Head Neck Surg 2007;133(8):772–775
20. Bailey CM, Clary RA, Pengilly A, Albert DM. Voice quality following laryngotracheal reconstruction. Int J Pediatr Otorhinolaryngol 1995;32(Suppl):S93–S95
21. Faust RA, Rahbar R. Robotic surgical technique for pediatric laryngotracheal reconstruction. Otolaryngol Clin North Am 2008;41(5):1045–1051, xi
22. Janssen LM, van Osch GJVM, Li JP, et al. Tracheal reconstruction: mucosal survival on porous titanium. Arch Otolaryngol Head Neck Surg 2009;135(5):472–478