Naiman AN, Ayari S, Froehlich P. Controlled Risk of Stenosis After Surgical Excision of Laryngeal Hemangioma. Arch Otolaryngol Head Neck Surg. 2003;129(12):1291-1295. doi:10.1001/archotol.129.12.1291
To evaluate the risk of subglottic stenosis after surgical excision of congenital subglottic hemangioma.
Retrospective analysis and case series.
Tertiary care teaching hospital.
A total of 13 pediatric patients diagnosed with subglottic hemangioma with unilateral, bilateral, or circular lesions and more than 50% airway obstruction between 1992 and 2001.
Open surgical excision was performed as a single-stage procedure either as primary or secondary intention. The cricoid cartilage was left open at the end of the procedure. Postoperative intubation was carried out in a pediatric intensive care unit.
Main Outcome Measure
An adequate airway after surgical excision.
All patients were successfully extubated. No recurrence was noted. Three patients developed subglottic stenosis, two grade 1 and one grade 2. All 3 showed a favorable outcome and did not require reintubation. One needed endoscopic management of the stenosis. Of these 3 cases, 2 occurred after carbon dioxide laser treatment (out of 3) and 1 after circumferential dissection (out of 3).
Extubation after surgery was successful in all cases of subglottic hemangioma. Risk of subglottic stenosis was limited and occurred only after circumferential dissection, especially if associated with prior traumatic laser damage of the hemangioma.
THE NATURAL history of hemangioma begins with a phase of rapid growth in the first few months of life (the reported mean age at diagnosis being 3.6 months),1 while the affected children develop progressive respiratory distress that may initially be intermittent then becomes continuous. On the other hand, hemangioma usually regresses by age 12 to 24 months, with complete involution within 5 years in 50% of cases and within 12 years in the other 50%.2 The impaired airway is managed either by bypassing the obstruction with a tracheotomy or by attempting to decrease the mass of the hemangioma. Thus, many therapeutic options have been proposed: radiation therapy, radioactive implantation, cryotherapy, electrocautery, sclerosants, carbon dioxide laser treatment, and surgical excision. Some of these have been gradually abandoned as ineffective or because of various severe immediate to long-term adverse effects.3
The most commonly accepted forms of therapy are currently steroids, carbon dioxide laser surgery, tracheotomy, interferon therapy, and open surgery. High doses of corticosteroids have been tried, but their effect on the airway obstruction was limited in many cases. Moreover, long-term systemic steroid administration causes such serious adverse effects as Cushing syndrome, growth restriction, and immune deficiency. These adverse effects can be minimized either by using small maintenance doses or by alternative-day therapy4 and can be avoided by using intralesional administration.5 The problems associated with prolonged intensive care, including long hospitalization and the financial aspect, may combine to favor the choice of intralesional steroids and intubation as treatment.
Tracheotomy alone is considered by many authors a standard treatment method, but associated complications, including a high risk of infection, accidental decannulation, tracheotomy tube plugging, and retarded speech and language development have been reported.6 Carbon dioxide laser surgery has proved its efficacy for small unilateral hemangiomas without extralaryngeal extension that do not require extensive or multiple applications. However, long-term follow-up has shown this therapy to be inadequate in cases of large hemangioma because of a significant risk of scar tissue formation.7
Several authors have reported cases of subglottic stenosis after carbon dioxide laser therapy.7,8 Potassium-titanyl-phosphate laser treatment has recently been advocated to manage subglottic hemangioma.9 All 8 patients treated by this method were successfully extubated or decannulated. This laser system seems to avoid the complications found with carbon dioxide laser, but these results have not been confirmed in larger studies. Massive cervicofacial hemangiomatosis with multiple airway sites, frequently associated with life-threatening complications, have been successfully treated using interferon alfa.10 In cases of extensive life-threatening hemangioma otherwise requiring tracheotomy, carbon dioxide laser treatment has proved ineffective but with a high-risk of stenosis.
Because of the unsatisfactory outcomes associated with available methods, it was obviously necessary to find other therapeutic options, and surgery has proved its efficacy and superiority in the management of such severe subglottic hemangioma. Initially described by Sharp11 in 1949, an external surgical approach for subglottic hemangioma was long considered a treatment of last resort, used only when the hemangioma had failed to regress by the time the child was 3 years old. Surgical excision was not a popular procedure for 2 main reasons: (1) an aggressive approach was not recommended in a benign tumor with spontaneous involution such as hemangioma, and (2) there was the risk of postoperative subglottic stenosis. In 1974, Evans and Todd12 described 3 cases of subglottic hemangioma treated with surgical excision using their original castellated laryngotracheoplasty technique, but tracheotomy and stenting were necessary. After improvement in pediatric intensive care techniques and progress in pediatric laryngeal surgery, single-stage procedures were developed.
In the last decade, several authors have used both techniques: classic laryngotracheal reconstruction13- 17 and single-stage procedures.14,16- 20 The superiority of single-stage laryngotracheal reconstruction over the classic procedure is clearly underlined by its avoidance of tracheotomy and stenting owing to the placement of a postoperative nasotracheal tube. The classic techniques have been replaced by single-stage procedures. The purpose of the present study is to assess the risk of postoperative subglottic stenosis after single-stage surgical removal of subglottic hemangioma.
From 1992 to 2001, 13 children diagnosed as having subglottic hemangioma (7 girls and 6 boys; aged 3 to 17 months; mean age, 6 months) were surgically treated in the Department of Otolaryngology–Head and Neck Surgery of the Edouard Herriot University Hospital, Lyon, France (Table 1). Surgical excision of hemangioma was performed in a single stage in all patients. The surgical approach involved an anterior horizontal incision facing the cricoid. The muscle strap was divided in the midline, and then laryngeal cartilage and the upper tracheal rings were identified. A vertical incision was made, encompassing the lower quarter of the thyroid cartilage, the cricoid cartilage, and 2 tracheal rings. A tube inserted through the lower part of the tracheal incision replaced the nasotracheal tube. When technically feasible, the nasotracheal tube was left in place. The submucosal dissection was started from the midline then carried out over the mass and finished between the lesion and the inner surface of the cricoid cartilage. The thyroid and tracheal incisions were closed, and the cricoid cartilage remained open at the end of the procedure. The skin was closed as for a cricoid split.
The patients were transferred to the pediatric intensive care unit under nasotracheal intubation. During intubation, antibiotics and antacid medication were administered. Systemic steroid therapy was started 24 hours prior to extubation and continued over 48 hours. Extubation was performend when the air-leak pressure in the endotracheal tube dropped below 20 cm H2O.
Postextubation endoscopy was performed only in cases of worsening stridor. Subglottic stenosis, when diagnosed, was assessed on the Cotton scale. The therapy required for the stenosis was determined by clinical and endoscopic assessment. Endoscopic treatment when necessary consisted of carbon dioxide laser surgery followed in the most recent cases by local mitomycin application. After treatment of the postoperative subglottic stenosis, if the child became asymptomatic, the outcome was considered favorable.
Open surgical excision of hemangioma was performed in all 13 cases (Table 1). The mean time between diagnosis and surgery was 4 months (range, 0-14 months), delay being due to other attempted treatments in 11 of the 13 children. All of these 11 children had received steroids, and 3 of them had developed cushingoid features. Three children were treated with carbon dioxide laser surgery.
The topography of the hemangioma and the obstructional grade of the subglottic area were established endoscopically (Table 1). In 4 cases, the hemangioma was bilateral, and in 3 cases circular. The median occlusion of the subglottic cross-sectional area was in all cases more than 50%, exceeding 75% in 10 patients.
After surgical excision, all patients were successfully extubated. The intubation time averaged 7.5 days, ranging from 6 days (8 cases) to 18 days (1 case). The patient extubated on the 18th postoperative day had undergone surgery after failure of carbon dioxide laser excision, and extubation had been attempted twice unsuccessfully (on the sixth and 12th postoperative days). Finally, the patient was accidentally extubated on the 18th day with good tolerance.
After extubation, 10 of the 13 patients were asymptomatic, and endoscopic control was not necessary. The mean follow-up for these 10 patients was 54 months, ranging from 13 to 116 months. No recurrence of hemangioma occurred during follow-up.
The patients previously treated by systemic steroids only (6 cases) and those without any prior treatment (2 cases) did not develop any severe postoperative complications. Only 1 of them, intubated for 9 days postoperatively, developed reactive granulation tissue on the anterior subglottic side, but endoscopic treatment was not necessary.
Postoperative symptomatic subglottic stenosis occurred in 3 cases. In 1 case, dissection was performed of a circumferential subglottic hemangioma with extension to the first tracheal ring (previously treated by 9-month systemic steroid therapy and short-term intubation). This patient developed circumferential stenosis in the inferior part of the subglottis (grade 1) but was not intubated. After 4 subsequent carbon dioxide laser applications, in the fifth postoperative month, the stenosis was considered stable and fibrous, causing 10% airway obstruction and remaining asymptomatic on follow-up (15 months). This patient received postoperative corticosteroid therapy, which was progressively decreased and stopped after 1 year.
The other 2 patients who developed postoperative subglottic stenosis had undergone previous laser procedures. The first showed a fine anterior grade 1 stenosis on the inferior subglottic side associated with granulation tissue in the anterior subglottis, which was excised endoscopically. The second child developed anterior glottic stenosis with anterior and lateral extension to the subglottis (grade 2). Neither of these cases of stenosis required endoscopic treatment. Corticosteroid therapy was stopped 4 and 7 months postoperatively, respectively. The stenosis decreased, became fibrous, stable, and asymptomatic during follow-up (at 8 and 24 months, respectively).
Single-stage laryngotracheoplasty has many proven advantages in the management of extensive subglottic hemangioma. Surgical excision offers a rapid solution for control of this life-threatening disease in a situation that, with other therapies, could require longer critical care. In our patients, a good airway was obtained after several days of intubation (on average, 7.5 days), and all children were successfully extubated. Postoperative evolution was favorable, with most (10/13) patients remaining asymptomatic throughout follow-up.
Age did not appear to be a limiting factor: in the present study, the patients initially treated by surgical excision were operated on at a mean age of 5.3 months. The same good results have been reported in the literature.14,16- 20 In 28 reported cases treated with single-stage laryngotracheoplasty, all patients were extubated after several days' intubation (2-15 days) with favorable evolution on follow-up (Table 2). The patients who underwent prior tracheotomy were successfully decannulated in all cases. Open surgery is a valuable method in the treatment of extensive subglottic hemangioma, allowing a sufficient airway to be obtained immediately after diagnosis and with favorable outcome on follow-up.
The main inconvenience of this method remains the tendency to develop postoperative subglottic stenosis. Three of our 13 patients developed grade 1 subglottic stenosis. Even so, 2 of them showed spontaneous favorable evolution. Only 1 case, a grade 1 circular subglottic stenosis, required secondary endoscopic treatment to achieve complete resolution. Of 43 open surgical excisions of hemangioma reported in the literature, 3 developed sublottic stenosis (Table 3). Occurring after excision of circular hemangiomas, 2 of these cases17 were of grade 1 subglottic stenosis and did not require additional therapy, having spontaneous favorable evolution. These postoperative stenoses were not severe, were easily manageable, and did not require tracheotomy.
Two complicating factors must be underlined in regard to the development of postoperative stenosis. The first is previous application of carbon dioxide laser. Two of the 3 patients in the present study who had undergone carbon dioxide laser vaporization as prior treatment developed subglottic stenosis. Submucosal dissection was much more damaging for the mucosa when surgical excision was performed after application of carbon dioxide laser: in one case it was necessary to sacrifice some lateral mucosa. In a similar case,16 the patient developed a severe stenosis and required tracheotomy to free the airway. This child, with a history of prolonged intubation (during aortic coarctation repair) and carbon dioxide laser as prior treatment, had bilateral hemangioma with extension to the second tracheal ring. Six weeks after surgical removal, the child developed a worsening stridor, and a tracheotomy was performed. This case represents the worst condition in which surgery can be performed: after carbon dioxide laser treatment and excision of an extensive subglottic and tracheal hemangioma.
The second complicating factor is circular hemangioma. One case of stenosis in our present series occurred after excision of a circular lesion previously treated by corticosteroid therapy and intubation. This was the patient who developed a circular grade 1 stenosis that required endoscopic treatment to stabilize the stenosis and free the airway. The remaining 2 cases of circular hemangioma did not develop stenosis and showed favorable postoperative evolution. In an earlier study in which 4 cases of circular hemangioma were treated with single-stage excision,17 2 patients developed grade 1 postoperative subglottic stenosis with favorable spontaneous evolution. Thus, it is clear that circular hemangioma involves only a limited risk of development of postoperative stenosis, and this does not amount to a contraindication for surgical excision.
Prevention of subglottic stenosis introduces the concept of cricoid enlargement as a preventive measure. This enlargement can be achieved by leaving the cricoid open, as in a cricoid split, or by placing a cartilage graft between the anterior edges of the cricoid incision. In all of our patients, cricoid enlargement was performed as in cricoid split.
In another series (14 cases managed by single-stage laryngotracheoplasty),17 anterior cartilage grafting was used, and 2 patients developed subglottic stenosis. There are not yet enough data available to assess the risk of stenosis with or without cartilage grafting.
Several recommendations for limiting the risk of subglottic stenosis may be made: (1) There is no need to wait for the hemangioma to grow. Open surgery is indicated immediately after diagnosis as primary intention rather than after failure of other therapies. (2) Carbon dioxide laser vaporization should not be attempted before surgery. (3) Cricoid enlargement prevents subglottic stenosis and should be performed at the end of the surgical procedure either by cricoid split or using cartilage graft. (4) Circular hemangioma is not a contraindication for surgery even though a limited risk of postoperative stenosis exists in this case, especially if carbon dioxide laser treatment has already been performed.
In conclusion, all of our children could be extubated after open surgical excision of subglottic hemangioma. Postoperative stenoses were not severe and proved easily manageable, with favorable outcome. The risk was highest after carbon dioxide laser and circular dissection of the subglottis. Cricoid enlargement prevented development of subglottic stenosis.
Corresponding author and reprints: Patrick Froehlich, MD, PhD, Otolaryngology Department, Hospital Edouard Herriot, Place d'Arsonval, 69003 Lyon, France (e-mail: firstname.lastname@example.org).
Submitted for publication June 20, 2002; final revision received March 25, 2003; accepted April 15, 2003.
This article was presented in part at the American Society of Pediatric Otolaryngology; May 14, 2002; Boca Raton, Fla.