Endoscopically placed expandable metal tracheal stents for the management of complicated tracheal stenosis☆

Materials and methods 

Six patients with severe tracheal stenosis treated between February 1998 and September 1999 are the subjects of this report. All patients had tracheal stenosis caused by endotracheal intubation and/or tracheotomy. All but 1 patient who presented with a long segment of tracheomalacia after prolonged intubation and tracheotomy had been previously operated for the tracheal stenosis. Four patients had undergone previous tracheal resection and primary reanastomosis, and 1 patient had undergone endoscopic dilatation. All 6 patients had associated pulmonary disease. In addition, 3 patients had confirmed gastroesophageal reflux disease (by esophagogastroduodenoscopy and biopsy and/or pH-metry) (Table 1).

Table 1. Clinical history

	Patient	Sex	Age (Yr)	Etiology	Previous Treatment (Time Before Stent Placement)	Comorbidities	Stenotic site (length) and % narrowing
	1	M	13	Intubation	Resection 2nd–5th TR and primary anastomosis (4 mo)	Bronchial asthma	Subglottis and anastomosis (5 cm)
					Dilation and intralesional steroid to anastomotic granulation tissue and stenosis (2 mo)		75%
	2	M	62	Intubation	Resection 1st–2nd TR and primary reanastomosis (6 mo)	COPD GER	Subglottis and anastomosis (3 cm)
					CO2 laser of web and anastomotic granulation tissue, intralesional steroid (5 mos)		80%
	3	F	74	Intubation Tracheotomy	Silastic stent (2 yr)	COPD GER, HTN	Site proximal to previous stent and long tracheomalacia (6 cm)
							80%
	4	M	67	Intubation	None	COPD, CAD, CHF, HTN, GER	Trachea (3 cm)
							80%
	5	M	60	Tracheotomy	Resection 3rd TR and primary reanastomosis	COPD	Anastomosis (2 cm)
							70%
	6	F	35	Intubation Tracheotomy	None	GER	At tracheotomy site (after decannulation)
						Bronchial asthma	Tracheomalacia (4 cm)

Abbreviations: M, male; F, female; COPD, chronic obstructive pulmonary disease; GER, gastroesophageal reflux; TR, tracheal rings; HTN, hypertension; CAD, coronary artery disease; CHF, congestive heart failure.

Operative procedure 

Under intravenous anesthesia, the patients are intubated through a Dedo laryngoscope using a 4.0 or 5.0 endotracheal tube (ETT). If the tracheal diameter does not allow the passage of the ETT, the stenosis can be rapidly dilated with a balloon dilator, and then the ETT is inserted. Proximal jet ventilation can also be used until the tracheal stenosis is dilated. Once the adequacy of the intubation is confirmed, the Dedo laryngoscope is suspended. Then, the endotracheal tube is removed and, under apnea, the stenosis is re-examined using a 4-mm, 0° rod lens telescope. The stricture length is calculated using the rod lens telescope. Likewise, the diameter of the stenotic and normal trachea is estimated comparing the lumen of the stricture with the diameter of a rigid bronchoscope or a rod lens telescope. The stenotic segment is then dilated with sequential bouginage or preferably with a balloon dilator of an adequate size (3-5 atmospheres for 1-2 minutes). Serial extubations and intubations are performed through the laryngoscope as needed (Fig 1).

View Large Image Download to PowerPoint

Fig. 1. (Clockwise from left top) Dilatation with balloon dilator and placement of metal stent under direct visualization.

The diameter of the stent should be slightly wider than that of the lumen of the normal trachea to assure a good fixation and should exceed the length of the stenotic segment by 10 mm to 20 mm. The stent is then placed and released (proximal or distal release) under visualization with the rigid telescope. The stent can be repositioned after partial release with ease; however, this is not the case after full deployment. Thus, the adequacy of placement is monitored using endoscopic visualization. Under 0° endoscopic visualization, the stent is further expanded against the tracheal wall using a balloon dilator. Using this technique with endoscopic visualization and repeated intubation/extubations via the suspended laryngoscope, a secure airway is maintained, direct visualization of stent placement is achieved, and fluoroscopy is unnecessary.

Results 

Our experience using expandable metal tracheal stents for the treatment of complicated stenosis over the past 48 months is summarized in Table 2.

Table 2. Stent characteristics, complications, and outcome

	Patient	Stent Diameter (mm) × Length (mm)	Complications	Additional Treatment by Months After Initial Stent Placement	Follow-up to Date After Stent Placement (mo)
	1	20 × 40 and 18 × 60	Granulation tissue and stenosis	2 mo dilation of stricture and YAG laser of granulation tissue proximal to stent	24
				8 mo stricture and granulation tissue requiring removal of stent, tracheal resection, and insertion of T-tube
	2	18 × 40	Granulation tissue	10 mo Dilation and excision (cold instrumentation) of granulation tissue	37
				15 mo Dilation and removal of granulation tissue (cold), re-expand stent
	3	18 × 70	Mild shortness of breath	None to date	22
	4	18 × 60	Restenosis	1 mo Dilation and placement of second stent 23 × 70 mm for stenosis proximal to stent	19 (dead of disease)
			Granulation tissue Stent fracture	6 mo Failed intubation attempt prompted removal of deformed and obstructing stent wires with associated granulation tissue
				12 mo Routine DLT found granulation tissue proximal to stent and narrowed stent to 75% obstruction; not dilated
				16 mo YAG laser ablation of granulation tissue and the removal of deformed and fractured stent wires
	5	20 × 40	Hemoptysis Granulation tissue	2 mo DLT and esophagoscopy for hemoptysis; no granulation tissue or stenosis; stent patent	26
				8 mo DLT for shortness of breath; granulation tissue at distal aspect of stent, stent re-expanded
	6	20 × 60	None	None	48

Initially, all patients enjoyed immediate palliation of the tracheal stenosis with resolution of the stridor, shortness of breath, and dyspnea on exertion. Long-term results, however, have not been as positive. Four patients developed significant granulation tissue and recurrent stenosis, requiring intervention within 6 months after placement of the stent (patient numbers1, 2, 4, and 5). One patient (number 4) presented with stenosis proximal to the stent 1 month after placement and required placement of a second stent overlapping the first one. Two patients (numbers 1 and 4) eventually required removal of the stent. In 1 patient (number 4), significant deformation and fracture of the stent was identified during attempts at intubation for other surgeries (Fig 2).

View Large Image Download to PowerPoint

Fig. 2. Granulation tissue and deformed wires causing significant obstruction of the airway.

Discussion 

The best approach for the management of extensive tracheal stenosis is yet to be defined. Tracheal stenosis involving more than 50% of the tracheal length, those associated with extrinsic or intrinsic malignant obstruction, or those combined with subglottic and/or glottic stenosis or tracheomalacia and tracheal stenosis recurrent after resection of 4 or more tracheal rings, offer the greatest challenge for management. Tracheoplasty using free grafts5 and vascularized flaps6, 7 and/or endoscopic resection of scar are often associated with a high incidence of complications and/or recurrence. Granulation tissue is a frequent cause of anastomotic obstruction11 after a segmental resection and primary end-to-end anastomosis of the trachea. Because of the lack of universal efficacy of the available techniques, in many patients with complicated tracheal stenosis, the goal of the treatment is palliation rather than cure.

Tracheal stenting has become an increasingly common approach to airway management as either as a temporary or permanent alternative or as an adjunct to open and endoscopic techniques. Stents provide resistance to scar contracture and provide support in areas of structural weakness because of cartilage loss. The ideal stent, however, should promote healing of de-epithelialized areas, by allowing mucosal ingrowth, should not impede airway mucociliary clearance, resisting bacterial contamination, and avoiding excessive pressure that would impede capillary circulation. There are many types of silicone stents, including the Montgomery T-tube (Boston Medical Products, Westborough, MA), Neville prosthesis, Westaby modification, Dumon, Hood, Nova, and Freitag stents.23 Silicone stents provide palliation for airway obstruction but interfere with the normal mucociliary clearance and may be obstructed with secretions.14 Other problems associated with silicone stents include migration, malposition, and granulation tissue formation. The Dumon stent is placed through a rigid bronchoscope using a stent introducer. After unloading, the stent is visualized with a telescope and manipulated with forceps into the best position.10, 12 The Montgomery T-tube (silicone) and Alboulker stents (Teflon, DuPont, Wilmington, DE) involve an external limb and/or require a tracheotomy. In our experience, the need for a tracheotomy has been the main reason for the patients' adamant refusal of a tracheal T-tube.

Several types of metallic stents have been used to treat tracheal stenosis.3, 10, 11, 12, 15, 16 These include the Palmaz (Johnson & Johnson, New Brunswick, NJ) and Gianturco (Cook Cardiology, Bloomington, IN) made of stainless steel, Wallstent (Boston Scientific Corp, Natick, MA) made of cobalt, and Strecker (Medi-Tech, Inc, Miami, FL) and Ultraflex (Boston Scientific Corp, Natick, MA) made with titanium or nitinol.23 Metal stents are introduced in a compressed form and released into the airway using rigid bronchoscopy and/or fluoroscopic guidance. Stents are expanded by balloon dilatation or by inherent recoil. Those that expand by inherent recoil have less risk of overexpansion and associated perforation while still assuring fixation against the tracheal wall. In our patients, we did not encounter any patient in whom the stent had migrated spontaneously. This has been corroborated by others.11, 16 In contrast to silicone stents, metallic mesh stents such as the Ultraflex, the Wallstent, and the Palmaz theoretically allow epithelization through the stent mesh maintaining near-normal mucociliary function. Other proffered advantages include incorporation into the tracheal wall, simple insertion, and high visibility on chest radiograph.10

The nickel-titanium alloy (Nitinol), expandable tracheal stents (Ultra-flex System; Boston Scientific) have been used successfully to palliate obstructing bronchial or tracheal tumors. According to some, Nitinol stents are well tolerated and have been used without untoward with reaction or delays in mucosalization.11, 17 Ducic and Khalafi11 reported the use of the Nitinol tracheal stent in the treatment of benign symptomatic stenosis of the cervical trachea. Complete mucosalization of the intraluminal surface of the stent was noted in each case during surveillance endoscopy performed 6 to 8 weeks postoperatively. Likewise, Remacle et al24 reported the treatment, complications, and follow-up of 23 patients with tracheal stenosis treated with a Gianturco stent (Cook Cardiology). In their series, they encountered only 3 patients with granulomas.

Despite initial success achieving resolution of airway symptoms, 4 of our 6 patients developed significant granulation tissue (patient numbers 1, 2, 4, and 5), requiring endoscopic removal of the granulation tissue and dilation within 6 months of their placement. Others have associated the use of metal stents with a high rate of granulation tissue and difficulty in removal or manipulation. Furman et al18 reported the use of the expandable metallic Palmaz stent (Johnson & Johnson) in the treatment of pediatric tracheomalacia and bronchomalacia encountering a high number of complications. Granulation tissue formation was the most common and serious problem. One of their patients required the removal of 2 tracheal stents because of granulation tissue and 2 patients died, 1 possibly related to the stent placement. DeRowe at al16 used a nickel-titanium spiral coil stent (InStent, Inc Medtronic, Eden Prairie, MN) in 5 adult patients with subglottic stenosis. Within 2 months, 1 patient developed granulation tissue inferior to the original stenotic segment that required removal of the stent and placement of a T-tube. The remaining 4 patients did well with more than 6 months follow-up and did not develop granulation tissue.

Recently, Casiano et al22 published their results using WallStent Endoprosthesis (Boston Scientific Corp) in 13 patients. They report that 10 of 13 patients remained stridor free after an average follow-up of 15 months. However, a detailed analysis of their data reveals that 8 out of 13 patients develop a restenosis (4 of these greater than 30% of lumen), and 6 out of 13 required additional surgery (granuloma, stent placement, or removal).

One can postulate several possibilities to explain the vast differences in outcome reported by others. Characteristics of the metal alloy may contribute to tracheal tissue reaction induced by sensitivity to certain metals or due to electrolysis of the metal. Gastroesophageal reflux disease may contribute further by accelerating electrolysis and by producing epithelial and subepithelial damage. Although we prescribed proton pump inhibition to all our patients, the issues of dose, frequency, and adequacy of response, as well as patient compliance could not be ascertained. Pressure necrosis caused by excessive expansion of the stent because of oversize or overenthusiastic expansion at the time of placement may induce the formation of granulation tissue. As previously mentioned, stents that expand by their inherent recoil are less prone to cause this complication. Similarly, characteristics of the patient such as healing, pressure of chondritis, and chronic pulmonary disease with overgrowth of bacteria in the presence may also contribute to the formation of granulation tissue. Although probably impossible to ascertain the precise role of all these factors, they should be taken into consideration when recommending stenting for chronic tracheal stenosis.

Conclusions 

Complicated tracheal stenosis is a difficult management problem, for which metal stents may provide temporary palliation. Our placement technique provides easy access to the airway for repeated intubations and for the evaluation of the stenotic segment and can be applied to other problems that require a transendoscopic management. Visualization with rod lens telescopes during placement and expansion of the stent allows real-time confirmation of its location. Fluoroscopy with its inherent radiation exposure to both patient and staff is rendered unnecessary.

Metal stents, however, are associated with a very high incidence of obstruction with granulation tissue. Their use should be limited to a very selected group of patients with a short life expectancy (because of other comorbidities) or patients who are not amenable to reconstructive techniques and that cannot tolerate or accept a tracheotomy. After stent placement, patients should be monitored frequently (every 6-8 weeks) for the formation of obstructing granulation tissue.