Goals and Assessment

The goal of any surgical procedure addressing laryngotracheal stenosis is to establish a satisfactory airway, and eventually, to decannulate the patient. In addition, every effort should be made to preserve other important laryngeal functions—phonation, airway protection, and glottic closure. Surgical procedures designed to improve the airway often compromise other laryngeal functions. The anticipated overall function of the larynx should be carefully discussed with the patient so that expectations are reasonable.

Several factors should be considered, including the location, dimensions, and quality (soft vs. fibrous) of stenosis, associated vocal fold motion impairment, and extent of functional impairment. Initial evaluation should involve a thorough history that includes details of the degree of subjective impairment (perceived by the patient) along with objective measures (e.g., observing the patient walking 200 feet, climbing a flight of stairs, pulmonary function testing).⁶ Pulmonary function testing is helpful in delineating the site of obstruction and severity of airway compromise (Fig. 71-1).

Figure 71-1. Flow-volume loop for assessing adequacy of the upper airway. A, Normal result shows expiratory flow, indicated by positive deflection, and inspiratory flow, indicated by negative deflection. B, Loop from a patient with bilateral vocal fold motion impairment shows variable extrathoracic obstruction with a mid–vital capacity inspiratory flow rate (Vi₅₀) of less than 1.5 L per second. C, Loop from a patient with an infiltrative tumor and a fixed obstruction. BTPS, body temperature, pressure saturation.

Physical examination, including an indirect laryngeal examination followed by direct laryngoscopy and bronchoscopy, is essential (Fig. 71-2). High-resolution computed tomography of the larynx and trachea may be useful in assessing the level and extent of the stenosis. Three-dimensional computed tomography with volume rendering or virtual endoscopy has been employed to evaluate the extent or severity of the stenosis.^7,8

Figure 71-2. Endoscopic view of subglottic stenosis for preoperative staging and evaluation.

An important consideration is whether the site of stenosis requires reestablishment of structural support, usually by repositioning of existing cartilage or, more commonly, through the use of cartilage or bone grafts.⁹ Even with grafts, the preservation of existing mucosa and the judicious use of antibiotics, stents, skin or mucosal grafts are all adjunctive treatments intended to minimize formation of granulation tissue, and subsequent deposition of collagen.

Timing of Repair

A chronic laryngotracheal stenosis repair is generally elective, once the appropriate workup is completed. The onset of symptoms may be insidious, depending on the extent of scar formation and contracture, and in intubated patients, several attempts at extubation usually fail, requiring tracheotomy.¹³ Such patients subsequently cannot be decannulated or, after a brief decannulation, experience gradually progressive upper airway symptoms. Irreversible stenosis is usually diagnosed at this time, and repair can be performed electively after thorough evaluation. A key consideration is that patients with inflammatory or autoimmune diseases require stabilization of their underlying disease processes before surgical management of the stenosis.

Endoscopic versus Open Repair

Beginning in the 1940s, the literature describes the use of endoscopic division of webs and subsequent keel placement. These early endoscopic techniques included scar removal via instruments or lasers followed by application of mitomycin-C. The evolution of powered instrumentation such as better microlaryngeal instrumentation and refinements in laryngoscopes have both enhanced visualization and maneuvering and have led to an increase in the numbers of laryngeal lesions removed via endoscopic techniques that previously required open approaches. Duration of inpatient hospital stay and morbidity of procedures have both declined with endoscopic techniques; however, the skill and comfort level of the surgeon with these newer tools cannot be overemphasized.

The last few decades have seen the growing popularity of the carbon dioxide (CO₂) laser, which offers the advantages of delayed formation and maturation of collagen in wounds, thus allowing reepithelialization before scar formation as well as minimal deep tissue injury. The laser has enabled precise control of the areas removed and hemostasis with the aim of preserving existing mucosa that may be utilized for the repair. Limitations of the CO₂ laser include the risk of an airway fire, and the danger of the laser plume for the surgeon and hospital staff if they are working with infectious lesions.¹⁰ In addition, it is often difficult to control thermal injury, which may result in perioperative edema and postoperative scarring of the vocal folds.¹¹ “Due to these risks,” some authors advocate initial endoscopic management of chronic laryngotracheal stenosis, reserving open repair for patients whose stenosis does not resolve with endoscopic management.¹² The advantage of this approach is that an open surgical repair may be avoided in some instances; however, several endoscopic procedures may be required for long-term success. Current recommendations include treatment of mild stenosis (Cotton-Myer grades I and II) with endoscopic techniques. Failure of initial endoscopic technique is associated with previous surgery, circumferential scarring, loss of cartilaginous support, exposure of cartilage leading to chondritis, posterior inlet scarring with subsequent arytenoid fixation, postoperative bacterial infection, and initial poor choice of candidates, such as patients with combined laryngeal and tracheal stenosis or vertical scar length greater than 1 to 1.5 cm. With prudent choice of candidates and careful postoperative care and monitoring, success of decannulation at 1 year after endoscopic surgery for stenosis has risen from 60% to 65% to 80% to 85%.

Postoperative management should include antibiotics for 1 to 3 weeks, antireflux management, and subsequent wound reassessment in 6 weeks via endoscopy. Patients with more severe stenosis may be better served with an initial open technique, although this does not guarantee higher success rates in such difficult cases.

Microsurgical Débridement

Much like the impact of advanced instrumentation for endoscopic techniques in laryngeal surgery, the success of powered instrumentation in many arthroscopic procedures has led to the popularity of microsurgical debriders in airway surgery. This technique combines the use of flexible forceps and a microsurgical debrider with videoendoscopic control.¹⁴ The debrider has a powered rotating blade and suction so that soft tissue lesions are aspirated into the tip of the blade and cleanly cut. Use of this instrumentation in endoscopic sinus surgery provided for more efficient and precise tissue removal and reduced injury to normal mucosa. Development of a laryngeal microsurgical shaver in the 1990s allowed for removal of bulky, exophytic laryngeal papilloma in patients with recurrent respiratory papillomatosis, obstructing carcinoma, and Reinke’s edema.^15–17 In comparison with the carbon dioxide laser, the laryngeal microsurgical debrider eliminates the risk of airway fire because there is no laser plume and therefore a lower infectious risk for the operating room staff. Anecdotal data show that use of this instrument in patients with recurrent respiratory papillomatosis, which was previously treated with the laser, have less postoperative pain and a quicker return to a usable speaking voice. A decrease in operative time has also been observed, thus reducing both the patient’s exposure to a general anesthetic and overall operating room costs.¹⁸

We use the microdebrider for many laryngeal lesions (Box 71-2) by maneuvering the rigid telescope with the laryngeal shaver, thus promoting camera steadiness and good visualization of the blade tip relative to the lesion as it is removed. Subglottic and tracheal stenotic lesions are typically removed with a surgical microdebrider (4-mm) subglottic blade to excise the fibrous scar. Softer, bulky lesions, such as laryngeal and tracheal papilloma, are rapidly removed with the 4.0-mm round window (skimmer) blade. Papillomas and other lesions involving the true vocal folds are removed with the 3.5-mm round window (skimmer) laryngeal blade (Figs. 71-3 and 71-4).

Figure 71-3. Powered instrumentation. Microdebrider blade (upper inset), for use in larynx, subglottis, and trachea, and skimmer blade (lower inset), for removal of mucosal lesions.

Figure 71-4. Endoscopic view of endolaryngeal lesion removal using the round blade of a microdebrider.

Local Application of Mitomycin C

Topical mitomycin C may be useful for the treatment and prevention of subsequent re-stenosis and scar formation in the larynx and trachea. Mitomycin C is an antineoplastic antibiotic that acts as an alkylating agent by inhibiting DNA and protein synthesis. It can inhibit cell division, protein synthesis, and fibroblast proliferation.¹⁹

The use of topical mitomycin C remains controversial, with one study demonstrating that after 15 months of mitomycin application, all patients had clinical improvement of symptoms without evidence of recurrence.¹⁹ Another study performed on the pediatric airway found no statistical difference between a single topical dose of mitomycin and isotonic sodium chloride after laryngotracheal reconstruction.²⁰ Animal studies have shown a statistically significant reduction in the rate of restenosis after surgical lysis with the use of mitomycin C.²¹

Surgical Planning for Repair

The first consideration for successful repair of acute or chronic laryngotracheal stenosis is the establishment of an intact, reasonably shaped skeletal framework to provide a scaffold for the airway. Acute, blunt injuries to the larynx, in which cartilage is fractured and displaced, are generally best managed with open reduction and stabilization of the fragments. In contrast, chronic injuries often are devoid of cartilage as a result of resorption of devitalized segments or of chondritis and subsequent chondromalacia or necrosis. The cartilage has often been replaced by scar tissue, so cartilage or bone grafts are necessary to reestablish structural support and to provide luminal augmentation. There is no consensus as to the ideal graft materials, and the grafts described in the literature use rib, iliac crest, hyoid bone, epiglottis, thyroid cartilage, composite auricular cartilage, and nasal septal cartilage.

Each of these graft materials may be limited by size, lack of pliability, and variable resorption due to lack of blood supply. The rib or costal cartilage graft is most commonly used because of its size, strength, and relative ease of harvesting. Vascularized hyoid and thyroid cartilage grafts have shown considerable success^22,23 but are limited by the amount of cartilage available. Composite auricular and nasal septal cartilage has the advantage of providing simultaneous epithelial repair, but it generally lacks adequate size and rigidity for cases in which large grafts are required. Vascularized perichondrial grafts and periosteal grafts have been evaluated for their chondrogenic or osteogenic capabilities and show potential for providing a pliable, vascularized graft that generates viable rigid bone or cartilage. In a rabbit model, vascularized perichondrium was shown to form significantly more cartilage when placed in a subglottic defect, and it tolerated a 2-hour ischemia time²⁴; it has also been successfully maneuvered into tube form for reconstruction of a segmental tracheal defect.²⁵ The sternocleidomastoid myoperiosteal flap has been successfully used for subglottic and tracheal reconstruction, with a low incidence of bony overgrowth on long-term follow-up.²⁶

Use of Skin and Mucosal Grafts

The second consideration of the successful repair of acute and chronic laryngotracheal stenosis is the establishment of a completely epithelialized lumen of reasonably normal size and shape. In acute injuries, the ideal method is primary closure of mucosal lacerations after minimal débridement of nonviable tissue. In chronic stenosis, scar tissue should be excised, with preservation of as much overlying mucosa as possible. Small mucosal defects can be left to remucosalize, resulting in minimal detrimental effects.²⁷ The controversy arises in the use of epithelial grafts to resurface large defects in acute and chronic stenosis. Wounds that heal by secondary intention form granulation tissue and deposit collagen, which may contract until later covered by epithelium.²⁷ When the denuded area is secondarily infected—particularly after tracheotomy—epithelialization can be prolonged, leading to excessive granulation tissue and scar formation.²⁸

Epidermal grafts have been advocated to interrupt this process and to result in healing similar to that of primary intention with minimal collagen deposition. When a stable cartilaginous framework is present, the tendency for the epidermal graft to contract is less, because the laryngeal skeleton acts as an external splint.²⁹ The techniques of skin grafting as an important adjunct in laryngotracheal reconstruction have been well described.³⁰

However, the disadvantage of epidermal grafts in laryngotracheal reconstruction is that the larynx is not an ideal site to accept free epidermal grafts. The larynx is in constant motion with swallowing and neck movements, and the bed is potentially contaminated when a tracheotomy is present. A nonvascularized epidermal graft in the airway can become infected, thereby exacerbating the process of healing by secondary intention. The benefit of an epidermal graft must be weighed against the possibility of infection, especially in the context of an underlying skeletal reconstruction.

Stent Placement

Internal laryngeal and subglottic stenting is widely used with acceptable results; however, questions about the indications for stenting, the optimal type of stent, and the duration of stenting are largely unanswered. Internal stenting with soft or firm materials increases the risk of local infection, mucosal ulceration, and granulation tissue, which is directly related to the duration of stenting. Capillary blood flow ceases with a pressure of 20 to 40 mm Hg in mucosa closely adherent to cartilage, possibly initiating ischemic mucosal injury.³¹ Therefore, several writers advocate avoiding internal stenting (unless absolutely necessary) and minimizing the duration of stenting. However, internal stenting is commonly used for the following specific indications: (1) to provide support for cartilage and bone grafts or to splint displaced cartilage fragments in the desired position, (2) to allow approximation and immobilization of epidermal grafts to a recipient site, (3) to separate opposing raw surfaces during healing, and (4) to maintain lumen in a reconstructed area that lacks adequate cartilaginous support and requires scar formation.

The optimal stent has not been determined, and a variety of options exist. A tantalum³² or umbrella silicone³³ keel is useful for anterior glottic webs, but for more severe laryngotracheal stenosis, various stents are commonly used, including the Montgomery laryngeal stent (Fig. 71-5), the Aboulker stent, the “Swiss roll” siliconized rubber stent, and the finger cot stent; each has proponents who describe good results.^34,35 In general, a soft stent should be used to minimize pressure on mucosal surfaces when cartilaginous support is satisfactory and when the indication for stenting is to splint an epidermal graft or to separate opposing raw mucosal surfaces. A finger cot with a povidone-iodine–soaked sponge is an excellent soft stent; an alternative is a Montgomery laryngeal stent, which is firm but conforms to endolaryngeal contours. A firm stent is needed when the cartilaginous framework or graft requires splinting or when inadequate cartilage is present, thereby requiring that luminal support be achieved through collagen deposition and scar contraction. A solid stent is preferable to a hollow stent to minimize aspiration. However, a soft, hollow stent such as the Montgomery T-tube allows phonation. The T-tube is commonly used for subglottic and upper tracheal stenosis; appropriate patient selection and education are essential, because this is a single-lumen tube with a greater risk of airway obstruction.³⁶ An Aboulker stent can be used for subglottic or combined subglottic and laryngeal stenosis. The advantage of this stent is that it is used with a double-cannula metal tracheostomy tube, and is secured to the tube, thus preventing removal. Modifications of this technique that allow tracheotomy changes have been described.³⁷

Figure 71-5. Thyrotomy with Montgomery stent in place.

There is no uniformly accepted optimal duration of stent placement. Most authorities recommend 6 to 8 weeks, but the duration varies from 2 weeks to several months.³⁸ Usually 1 to 3 weeks is sufficient if only mucosal healing or epidermal graft take is required. If the laryngeal skeleton is adequate but requires splinting to maintain the appropriate position, 6 to 8 weeks is the generally accepted time frame.

However, if the cartilaginous framework is deficient and mature scar is important for structural strength, prolonged stenting is indicated to allow scar contraction around an inert, firm object. Successful prolonged stenting has been reported for intervals of up to 14 months. In this situation, the problem arises in premature stent removal, usually before sufficient scar contraction has occurred.³⁸ The presence of a paratubal air envelope, which manifests as linear air densities between the outside wall of the stent and the inside wall of the larynx and trachea, is reported to be a reliable indicator of laryngotracheal structural integrity and functional stability and is a useful measure to determine optimal timing for stent removal.³⁹

Supraglottic Stenosis

Chronic hypopharyngeal stenosis usually results from unrecognized acute blunt injury to the hyoid and thyrohyoid membrane. The direction of force is posterior to superior and results in the following discrete injuries: (1) the epiglottis can be adherent to the posterior or lateral hypopharyngeal walls, (2) the hyoid bone can be fractured and displaced posteriorly with the epiglottis, thereby producing laryngeal inlet stenosis, (3) a horizontal web of the posterior hypopharyngeal wall may form at the level of the superior aspect of the epiglottis, or (4) postcricoid stenosis may result in stenosis of the esophageal introitus.⁴⁰

The surgical approach is through a transhyoid pharyngotomy. Tracheotomy is essential if it has not been previously performed. A horizontal skin incision is made over the hyoid bone, which is then identified. If the hyoid is fractured, attempts are made to reduce and wire it into its normal anatomic position; otherwise, the body of the hyoid can be excised. The vallecula is entered, and the mucosal incision can be extended to a lateral pharyngotomy, if needed, with care taken to preserve the superior laryngeal neurovascular pedicles and hypoglossal nerve. After satisfactory exposure has been provided, attention is directed to repair of the stenosis.

Adhesions of the epiglottis to the hypopharyngeal walls may be managed by division of the adhesion along its long axis, submucosal excision of scar tissue, and primary mucosal closure. When a horizontal web or shelf is present, a vertical incision is made through the web, and scar tissue is excised. Mucosal flaps are undermined to allow closure of the incision in a horizontal line. If necessary, superiorly based mucosal advancement flaps can be used. If the hypopharyngeal stenosis is extensive, the division of bands and the excision of scar tissue may result in an extensive denuded area over the constrictor muscle, and skin grafting should be considered. Hollow silicone esophageal stents have also been used to assist in the healing of a skin graft.⁴⁰ If resection of scar tissue results in an extensive full-thickness pharyngeal defect, a radial forearm revascularized flap may be used to provide thin, pliable, soft tissue, which may, however, result in an adynamic hypopharyngeal segment.

Laryngeal inlet stenosis results from posterior displacement of the hyoid and epiglottic cartilage and can be accompanied by a fracture through the thyroid notch. This injury is approached through a laryngofissure approach with division of the thyrohyoid membrane and elevation of the mucoperichondrium on the laryngeal surface to allow for excision of this inverted wedge of the base of the epiglottic cartilage, perichondrium, fascia, and scar. With suturing of the resultant flaps to the free edge of the anterior epiglottic perichondrium, the thyrotomy is closed, and stenting is not required.⁴⁰

Glottic Stenosis

Posterior Glottic Stenosis

Posterior glottic stenosis is most commonly caused by endotracheal intubation, although mechanical fixation may also occur from cricoarytenoid joint arthritis. Posterior glottic stenosis may be confused clinically with bilateral vocal fold paralysis before direct laryngoscopy. It should be considered when the history is appropriate and when there are indirect laryngoscopy findings of anteromedial displacement of the arytenoid cartilages, which thus obscure the posterior commissure scar, with a pseudo-foreshortening of the anteroposterior length of the membranous folds or glottic aperture.

Mechanical fixation and neurogenic paralysis are differentiated by direct laryngoscopy, and palpation of the arytenoid to assess mobility is essential. Joint excursion should be uninhibited. Passive medial movement of the contralateral arytenoid when the ipsilateral arytenoid is displaced laterally suggests interarytenoid scar. Electromyography findings of denervation or action potentials of the vocalis muscle are very useful to document neurologic integrity, particularly when results of direct observation of the vocal fold movement are equivocal. Patients with such findings may have a preexisting tracheotomy, which decreases posterior cricoarytenoid muscle activity, thereby making active lateral vocal fold excursion difficult to visualize unless the tracheotomy is plugged temporarily.

The extent of injury is variable. Type III and type IV posterior glottic stenoses are difficult to manage successfully, particularly when they are associated with subglottic stenosis.^3,43 The major therapeutic challenges are to reestablish a satisfactory airway and to preserve the voice, which is often normal.

Endoscopic repair of posterior glottic stenosis is successful in carefully selected cases. Patients with an interarytenoid web and posterior sinus tract have been managed with a simple division of the web. Many patients experience restenosis, particularly those who have had a tracheotomy.⁴⁴ A soft, sponge-filled finger cot stent has been advocated in this situation to separate the raw opposing surfaces; it can be placed endoscopically and remain for 2 weeks.⁴⁵

Posterior commissure stenosis can be managed endoscopically with a CO₂ laser or a laryngofissure approach. Creation of a posterior micro-trapdoor flap is successful when a posterior sinus or a 3- to 4-mm posterior interarytenoid scar is present.⁴⁴ This technique fails when there is arytenoid fixation or a scar that is more than 1 cm in height. When posterior interarytenoid mucosa is insufficient to create a trapdoor flap (>3-4 mm), interarytenoid scar tissue is resected, with the mucosa preserved through a hockey-stick incision.

A synthetic implantation fabric keel may be placed endoscopically and left in place for 4 to 6 weeks to cause sufficient widening of the interarytenoid mucosa to allow for a micro-trapdoor flap. However, Dedo and Sooy,⁴⁴ who first described this procedure, reported that failures occurred in all of their patients who had arytenoid fixation. Several other writers recommend open repair through a laryngofissure approach.^45,46 The surgeon effects such a repair by developing a superiorly based mucosal flap, excising scar tissue and fibrosed interarytenoid muscle, and repositioning the mucosal flap to provide immediate epithelial coverage. Internal stenting is not required when both arytenoids are mobile. Care is taken to not enter the cricoarytenoid joints if both arytenoids are mobile. Successful surgical repair is an elusive goal.

Posterior glottic stenosis with arytenoid fixation has been traditionally managed through an external approach. If, after development of a mucosal flap and resection of scar, one arytenoid is freely mobile, resection of the fixed arytenoid is unnecessary. When bilateral arytenoid fixation exists, removal of the less mobile arytenoid is necessary to achieve a satisfactory airway. Denuded mucosal surfaces are covered with mucosal flaps, skin, or mucosal grafts. Stenting for 2 to 3 weeks with a conforming silicone laryngeal stent or with a soft finger cot stent is required. A posterior cricoid split with rib graft may also be considered (Fig. 71-6).

Figure 71-6. Rib graft for posterior subglottic stenosis. A, Schematic of anterior wall defect and graft placement. B, Anterior wall defect (intraoperative photo). C, Rib graft. D, Rib graft in place.

In 1983, Ossoff and associates⁴⁷ first described complete arytenoidectomy via an endoscopic approach. Endoscopic laser arytenoidectomy is also useful for managing type IV posterior commissure stenosis when subglottic involvement is limited, because this procedure avoids the complications of an external approach and allows for the immediate assessment of airway size (Fig. 71-7). Endoscopic laser arytenoidectomy requires the removal of the entire arytenoid except for the muscular process. Arytenoid perichondritis is a potential complication, and the administration of perioperative antibiotics and antireflux measures, as always, are recommended. The degree of vocal fold lateralization obtained is variable, and the use of a lateralization suture or a finger cot stent may improve the final result by controlling the extent of lateralization.

Figure 71-7. Endoscopic view of an arytenoidectomy.

The external approaches for an arytenoidectomy are through a laryngofissure or a Woodman arytenoidectomy. The latter approach exposes the arytenoid cartilage through a posterolateral extralaryngeal approach and the entire arytenoid cartilage is resected, except for the vocal process. A submucosal suture is placed through the vocal process and anchored around the inferior thyroid cornu to lateralize the vocal fold (6 mm in men, 5 mm in women).⁴⁸ Other writers believe that tethering the vocal process around the inferior thyroid cornu may reposition the vocal fold inferior to the opposite fold; they recommend suturing to the thyroid lamina at the vocal fold level⁵ or around the superior cornu (success rate, 60%-80%).⁴³ If the procedure is unsuccessful, completion arytenoidectomy through a laryngofissure approach with suture lateralization of the vocal fold can be performed.⁴⁹

Exploration of the fixed cricoarytenoid joints with lysis of adhesions without arytenoidectomy has also been advocated. Fibrous bands on the medial aspect of the cricoarytenoid joints are divided, and the arytenoids are stented in full abduction for 2 to 3 weeks; stent removal and aggressive speech therapy follow. Return of vocal cord mobility has been described in 90% of patients managed in this way.⁵⁰

Complete Glottic Stenosis

Complete glottic stenosis usually results from unrecognized severe extralaryngeal trauma, in which injury is rarely isolated to the glottis and associated subglottic injury is common. Endoscopic management is rarely indicated because of the severity and nature of the injury, and only a few instances have been reported in the literature. The mainstay of management involves an open laryngofissure approach. The stenosis should be divided at the midline and excessive scar tissue resected; the mucosa is preserved and mucosal flaps are developed from the aryepiglottic folds for coverage. If extensive areas are devoid of mucosa, epithelial grafts of buccal mucosa, grafts of nasal septal mucosa, or split-thickness skin grafts may be used. Large split-thickness skin grafts may cause excessive crusting. The grafts are sutured in place and stented with a form-fitting laryngeal stent for 4 to 8 weeks.⁵¹ When the stent is removed, an umbrella keel is placed into the anterior commissure for 2 weeks.

An alternative approach is reconstruction with an epiglottic flap. This approach is indicated in the following situations: severe glottic stenosis, a more than 50% reduction in the anteroposterior dimension of the glottis, glottic-subglottic stenosis, and glottic-supraglottic stenosis with an intact epiglottis.^52,53 The repair is approached through a midline thyrotomy, and the thick scar is transected in the midline. Submucosal excision of scar tissue can be attempted, and the mucosa can be repositioned and secured with chromic suture. The base of the epiglottis is identified, and the median thyroepiglottic ligament is transected. The epiglottis is pulled inferiorly to reach the anterior cricoid arch, and the epiglottic flap is sutured to the outer anterior edges of the thyroid cartilage laterally and to the cricoid cartilage inferiorly. This procedure results in an epithelialized, widened anterior commissure. Several cartilage-splitting incisions may be used to allow the petiole to be folded onto itself to make a sharper anterior commissure.³⁶

Subglottic Stenosis

Endoscopic management of subglottic stenosis requires careful patient selection. The predictive factors for failure were described earlier.^43,54,55 CO₂ laser excision followed by repair with a micro-trapdoor flap is moderately reliable for managing circumferential subglottic stenosis; it requires laser incision of the superior surface of the scar and the submucosal excision of scar tissue. An inferiorly based rectangular mucosal flap is created by two lateral knife incisions. The preserved flap does not completely cover the lasered area, but it does divide the raw concave area into two flat lateral raw areas, which heal rapidly. One study reported a 90% success rate in all patients undergoing the procedure, in whom scars that were less than 10 mm in height.⁴³ A large flap that can potentially obstruct the airway through a ball-valve effect should not be created. Other writers⁴⁴ have confirmed the efficacy of this procedure in adults and emphasize that it is useful when the stenosis is less than 10 mm in height, but a tracheotomy should be performed if the flap is longer than 4 mm. The procedure is uniformly unsuccessful in children, however.⁵⁵ Failure also occurs if associated posterior glottic stenosis or arytenoid fixation is present. Shapshay and coworkers proposed an alternative method, consisting of radial incisions followed by bronchoscopic dilation.⁵⁶ Their series of patients was small, however, and follow-up observation was only 1 year; thus, success was not assumed.

Subglottic and tracheal stenotic lesions may be microsurgially debrided with use of a microdebrider (4-mm) subglottic blade. External repair is indicated when endoscopic techniques fail or when the extent of the stenosis is severe and factors are unfavorable for an endoscopic approach. Various techniques are available, and the selection of the appropriate repair technique should anticipate the amount of rigid support required for luminal augmentation and the possibility that scar resection requires epithelial grafting. Subglottic stenosis due to extensive mature scar with adequate cartilaginous support can be repaired via a laryngofissure approach.

Subglottic scar is resected to cricoid perichondrium and is relined with the buccal mucosa or a split-thickness skin graft. The graft is preferably sewn around a stent that is fixed in place. When cartilaginous support or luminal augmentation is also required, various autogenous cartilage grafts are available. The hyoid-sternohyoid muscle interposition graft has been extensively used, with a success rate of approximately 70% in adults.⁵⁷ This technique can be applied to isolated subglottic stenosis or subglottic stenosis in combination with anterior glottic or tracheal stenosis.

The procedure uses a vascularized segment of the body of the hyoid pedicle on the sternohyoid muscle and its adjacent periosteum. The length of bone can be tailored to the length of the stenosis; it is interposed vertically into the stenosis and secured with four-point fixation. A stent is not used unless an epidermal graft is required after scar excision to minimize possible graft resorption by pressure necrosis.

The advantages of this technique include immediate reconstruction with a single operative field, theoretically better tissue survival with minimal graft remodeling, and versatility in augmenting glottic, subglottic, and upper tracheal stenosis. The long-term survival of the bone graft has not been conclusively shown; the graft is limited in the amount of augmentation it provides because of its narrow width and concave shape. Additional procedures are often required to remove granulation tissue. The thyroid-cartilage-sternothyroid pedicle composite graft was developed to overcome the limitation of the hyoid-sternohyoid graft. A large central piece of thyroid ala attached to the sternothyroid muscle through its perichondrium is harvested. A flap of contralateral thyroid alar perichondrium is harvested in continuity with the ipsilateral perichondrium and serves as internal lining when the composite graft is transposed. Stenting for 1 to 6 weeks is required. This technique also offers the advantages of the hyoid-sternohyoid graft, except that it cannot be used for glottic stenosis. A larger piece of cartilage, however, is available to reconstruct more extensive defects. The internal perichondrial transposition allows for remucosalization with minimal granulation tissue formation.

Various free autogenous cartilage grafts have been used for luminal augmentation. These grafts have variable amounts of resorption and require prolonged stenting to allow resorbed graft to be replaced by firm, mature scar tissue. Costal cartilage grafts have the advantage of providing large grafts that are easily sculpted. Specific details of graft tailoring have been well described.⁵⁵ Composite nasal septal cartilage grafts have been used successfully in patients with stenosis of the larynx and upper trachea.¹⁵ Stenosis of up to 3 cm in height can be repaired, and the respiratory epithelium is a theoretic advantage, because it provides immediate resurfacing with a match of epithelium from the stenotic area. In children, a stent is not used, and the graft is not used so as to prevent injury to nasal growth centers. Various other autogenous grafts, including auricular cartilage, clavicular bone, and free hyoid bone, have been used but they have not received uniform acceptance.

Division of the posterior cricoid lamina should be considered when severe subglottic stenosis is present with posterior glottic stenosis or for complete glottic and subglottic stenosis. It is approached through an anterior cricotracheal fissure, avoiding disruption of the anterior commissure. The entire posterior cricoid lamina is divided vertically in the midline to the level of the postcricoid submucosa. Scar tissue excision is not necessary, and the interarytenoid muscle is divided if it is fibrosed. Prolonged rigid internal stenting is required for at least 3 months. Interposition cartilage grafts with perichondrium can be harvested from the thyroid lamina for posterior support, and anterior luminal augmentation can be performed.

Partial cricoid resection with thyrotracheal anastomosis is considered when other methods have failed. Several investigators reported good results and specific technical details. This procedure should be performed only in adults and is limited to stenosis involving the cricoid and upper trachea. There should be approximately 1 cm of normal lumen below the glottis. The main risks of this procedure are injury to the recurrent laryngeal nerves and anastomotic dehiscence with restenosis.

The procedure involves exposure of the involved laryngotracheal section with preservation of the recurrent laryngeal nerves. Identification of the nerves is often risky because of extensive scarring; therefore, dissection of the trachea is performed in the subperichondrial plane. The inferior resection line is immediately below the stenosis, beveling the anterior tracheal wall. The superior limit can include the cricoid up to the posterior lamina just below the cricothyroid joint. Care is taken to not injure the recurrent laryngeal nerves, which pass immediately posterior to the cricothyroid joints. The tracheotomy can be lowered if three of four tracheal rings can be interposed between the anastomosis and the tracheotomy; otherwise, the patient should be intubated and the anastomosis accomplished with 3-0 interrupted sutures placed submucosally and tied extraluminally. Laryngeal release procedures are usually necessary, and the neck is kept in flexion for 7 to 10 days postoperatively to minimize tension on the anastomosis. If a tracheotomy is not used, the patient is extubated in the operating room in 3 to 4 days.

Grillo and others^58,59 refined this procedure to incorporate a posterior membranous tracheal wall flap to resurface the bared cricoid cartilage. In a series of 80 patients with subglottic and upper tracheal stenosis, these researchers reported failure in only 2 out of 77 patients with at least 6 months follow-up; one perioperative death occurred. Of 77 patients, 66 had no restriction of daily activities due to dyspnea, and tracheotomy was rarely required.

A sliding flap tracheoplasty has been described, which is a variation of the cricoid resection and thyrotracheal anastomosis. The procedure involves resection of the anterior cricoid arch and the first two tracheal rings, with development of an inferiorly based anterior tracheal wall flap from the third through sixth rings; this is advanced and anastomosed to the thyroid cartilage superiorly. Patients for whom this procedure is selected should have a subglottic stenosis involving only the anterior cricoid arch and the first ring, an intact posterior wall, mobile vocal folds, and an uninvolved segment of mucosa below the glottis.

Cervical Tracheal Stenosis

Cervical tracheal stenosis in adults most commonly results from trauma produced by either intubation or tracheotomy, or blunt external trauma to the neck. Other causes are benign and malignant neoplasms, inflammatory diseases, and systemic autoimmune disease.

When surgical repair of tracheal stenosis is considered (Box 71-3), regardless of the cause, it is essential to determine the location, length, composition, extent of airway stenosis, and neurologic integrity of the larynx through indirect laryngeal examination, bronchoscopy, and radiography. In most instances, cervical tracheal stenosis can be categorized as cicatricial membranous stenosis, anterior wall collapse, or complete stenosis. Airway management should be considered preoperatively and discussed with the anesthesiologist, as for all of these procedures. If a tracheotomy is not present or is not planned, it is safest to have the patient breathe spontaneously until the airway is secured. If a Montgomery T-tube is used as part of the reconstruction, the patient should breathe spontaneously after the T-tube is inserted (Fig. 71-8). If the airway is 5 mm or smaller at the time of reconstruction, dilation is performed initially with use of general anesthesia; microsurgical debridement resection may also be performed. If the lumen is larger than 6 mm in diameter, intubation is carried out above the lesion, and dissection is performed carefully to prevent airway obstruction. If a postoperative tracheotomy is planned, it should be performed with use of local anesthesia at the outset of the reconstruction procedure.

Figure 71-8. A, Silicone T-tube stents for tracheal stenosis. B, Open placement of a T-tube.

Cicatricial Membranous Stenosis

Granular or fibrous stenosis of the cervical trachea with intact cartilage can initially be managed endoscopically. CO₂ laser excision of granular stenosis is useful because tracheotomy can be avoided, and the granular stenosis can be resected accurately with minimal bleeding. If the stenosis is circumferential, partial resections should be staged 2 to 4 weeks apart to prevent the formation of a circumferential denuded area of trachea in which restenosis can occur. Fibrous stenosis can be excised with a micro-trapdoor technique.⁴⁴

Open repair of a cicatricial membranous stenosis requires approaching the stenosis via a vertical midline incision through the anterior tracheal wall to expose the entire length of the stenosis. Care is taken to prevent lateral dissection, because the vascular supply to the cervical trachea from the inferior thyroid artery enters laterally. Submucosal excision of scar tissue is accomplished under direct visualization, and the tracheotomy may be moved to below the stenosis. An epidermal graft is useful to cover the denuded area, and it is usually fixed over a stent. Luminal augmentation is often required, and the type of graft used depends on the size of the graft needed and the surgeon’s experience and preference. The pedicled thyroid-cartilage-sternothyroid graft or a composite septal cartilage–mucoperiosteal graft seems to be reasonably successful.

Anterior Wall Collapse

Collapse of the anterior laryngeal wall can result from anterior blunt neck trauma, but it more commonly results from tracheotomy. Montgomery categorizes these injuries as suprastomal, stomal, or infrastomal.³² In this type of stenosis, a well-mucosalized posterolateral wall and residual cartilage in the area of the stenosis should be present.³⁶

A suprastomal and stomal stenosis can be approached and managed as described for cicatricial membranous stenosis. However, restoration of the anterior wall is essential. The surgeon can accomplished this restoration by (1) stenting the lumen open with a T-tube or a firm stent of a tracheostomy tube, such as the Aboulker stent, and repairing the anterior wall with a cartilage graft or (2) mobilizing the sternohyoid muscles and securing them together over the defect. In both instances, prolonged stenting is required.³⁶

Wedge resection of an anterior tracheal wall stenosis can be considered if (1) the stenosis is limited to two or three rings, with significant loss of cartilage that would preclude stenting, (2) the posterior wall mucosa is intact, or (3) the stenosis is stomal or suprastomal. The procedure involves adequate exposure of the involved trachea, including at least one ring above and one ring below the stenosis. Subperichondrial dissection is carried out carefully around the involved trachea to prevent injury to the recurrent laryngeal nerves. The stenosis is resected, with care taken to preserve posterior tracheal wall mucosa. An oroendotracheal tube is passed through the anastomosis, and the cartilaginous trachea is reanastomosed with submucosal sutures tied extraluminally. The patient is extubated when awake and alert, and granulation tissue is excised endoscopically as early as possible.³⁶

Complete Tracheal Stenosis

Segmental resection and primary anastomosis provide optimal results for complete tracheal stenosis. The average length of the adult trachea is 11 cm (range, 10-13 cm), and there are 14 to 20 tracheal rings. With hyperextension of the neck, 50% of the trachea is cervical. Approximately 50% of the trachea (5-7 cm) can be safely resected and anastomosed primarily with appropriate mobilization techniques. Other factors also should be considered in the determination of how much trachea can be resected. Older patients with calcification between tracheal rings may have less tracheal elasticity; a patient with a short, thick neck may have a deep trachea, with the cricoid at the level of the sternal notch. Tracheal mobilization is limited in a patient with decreased ability to extend the neck. The procedure is essentially the same as a wedge resection, except that dissection requires excision of the posterior membranous wall. The anterior cricoid arch and the posterior cricoid lamina are resected to just below the cricothyroid joints. An oroendotracheal tube is passed distal to the anastomotic site, and the anastomosis is performed with submucosal sutures placed around a trachea ring and tied extraluminally. Adequate mobilization of the larynx and trachea is required to relieve tension from the anastomosis. The neck is placed in extreme flexion, and a stay stitch is placed from chin to sternum.³⁶

Segmental resection with primary anastomosis shows promise as the most efficacious procedure for managing near complete or complete tracheal stenosis (Fig. 71-9). In patients in whom stenosis is limited to the trachea and subglottis and the glottis is normal, success rates in several large series are 95% to 97%.⁵⁹ When the stenosis also involves the glottis, resection with end-to-end anastomosis has been performed simultaneously with laryngofissure and decannulation repair of the glottic stenosis, which is achieved in approximately 95% of patients. In these instances, a Montgomery T-tube and prolonged stenting were uniformly used.

Figure 71-9. A, Near-complete tracheal stenosis. B, Segmental resection with primary anastomosis.

Extensive cervical and laryngotracheal stenosis not amenable to segmental resection and anastomosis can be managed with a multistaged trough reconstruction technique. The essential steps require the creation of a large tracheostoma by sewing cervical neck skin to residual tracheal mucosa. The second stage requires the formation of a rigid anterior wall by the insertion of Marlex mesh under adjacent cervical skin. Finally, the closure is obtained by tubing skin and Marlex mesh anteriorly to close the stoma and then advancing cervical neck skin as a second layer.

Laryngeal Release Procedures

Tracheal mobilization and laryngeal release techniques are required when a gap greater than 3 cm in tracheal continuity occurs in young patients. Older patients lose elasticity of the annular ligaments of the trachea and therefore require release procedures when more than 1 to 2 cm is resected. Three techniques can gain extra length for end-to-end anastomosis. The first is incision of the annular ligaments, which can add up to 2.5 cm but is most useful in younger patients. Incisions should be placed on one side of the lateral trachea above the anastomosis and on the opposite side below the anastomosis to preserve the blood supply to both tracheal segments. The second technique is superior mobilization of the distal trachea from the thorax, which can result in the achievement of 6 cm of mobilization. This technique requires transection of the left mainstem bronchus and end-to-side anastomosis with the right mainstem bronchus, and it carries a risk of mediastinal dissection. The third technique involves laryngeal release procedures, which can gain 5 cm of tracheal mobilization. The infrahyoid release involves transecting the sternohyoid, omohyoid, and thyrohyoid muscles at the level of the superior border of the thyroid cartilage. The greater cornua of the thyroid cartilage are divided bilaterally, and the thyrohyoid membrane is sectioned. This procedure carries a risk of injury to the superior laryngeal nerve and is associated with prolonged dysphagia.

Suprahyoid laryngeal release is often preferred. This technique requires transection of the insertion of the mylohyoid, geniohyoid, and genioglossus muscles to expose the preepiglottic space. The stylohyoid insertions are transected, and the body of the hyoid is transected medial to the insertion of the digastric tendon. This procedure allows the body of the hyoid, the thyroid cartilage, the cricoid cartilage, and the proximal trachea to drop inferiorly. Flexion of the neck affords an additional 1 to 2 cm of length.