Airway Simulation to Guide Stent Placement for Tracheobronchial Obstruction in Lung Cancer. Material and Methods

Transcription

1 Airway Simulation to Guide Stent Placement for Tracheobronchial Obstruction in Lung Cancer Joseph B. Zwischenberger, MD, Gerhard R. Wittich, MD, Eric vansonnenberg, MD, Raleigh F. Johnson, PhD, Scott K. Alpard, MD, Sanjay K. Anand, MD, and Robert J. Morrison, MD Division of Cardiothoracic Surgery, Department of Surgery and Departments of Radiology, Internal Medicine, and Anesthesiology, The University of Texas Medical Branch, Galveston, Texas Background. To effectively palliate large airway obstruction in advanced unresectable lung cancer (stage IIIB or IV), we developed an airway imaging technique to guide selective endobronchial metallic stent placement. Methods. Fourteen consecutive patients with severe dyspnea (American Thoracic Society grade 4) had a combination of fiberoptic bronchoscopy, chest roentgenography, computed tomographic scanning, helical computed tomography with three-dimensional reconstruction, and intraluminal bronchography with selective bronchial guidewire placement under fluoroscopy to visually reconstruct and simulate the abnormal airway before and during stent placement. Wallstent or Gianturco intraluminal stents were used alone or in combination (up to five stents) to establish patency of the distal trachea and the major bronchi. Results. All 14 patients had successful deployment with initial relief of airway stenosis (>75% predicted diameter). No procedural complications were noted. However, technical problems included stent foreshortening and imprecision of placement, misinterpretation of bronchography (mucous versus tumor), and airway maintenance during manipulation. Length of stay attributable to the procedure averaged 4 days. Stent placement initially improved the dyspnea score in 7 of 14 patients. Five of 14 died in less than 1 month, with the remainder alive at up to 8 months follow-up. Of those surviving more than 1 month, the Karnofsky score improved in 4 and was unchanged in 5, with 2 dependent (Karnofsky score <50), 3 functional (Karnofsky score, 50 to 70), and 4 active (Karnofsky score >70). Conclusions. A protocol combining helical computed tomography with three-dimensional reconstruction, bronchography, and bronchoscopy allows accurate assessment of malignant airway obstruction to facilitate intralumenal stent placement for relief of stenosis. Patient selection to favor effective palliation and cost effectiveness has yet to be defined. (Ann Thorac Surg 1997;64: ) 1997 by The Society of Thoracic Surgeons Initially designed for intravascular application, selfexpandable metal endoprostheses and delivery systems have been modified for use in the tracheobronchial tree. Previous studies have demonstrated the feasibility of placing tracheal or bronchial stents for benign or malignant lesions causing airway stenosis [1 7]; however, the total number of reported patients is small. Few studies have quantified the level of improvement in dyspnea [1, 3, 5, 7] or assessed stent placement efficacy in the treatment of complete airway occlusion in the absence of laser or bronchoscopic debulking. With complete or near-complete airway occlusion, the standard investigative modalities of bronchoscopy and computed tomographic (CT) scanning may give inadequate information to guide decision making regarding the amount of viable lung distal to the obstruction or the technical feasibility of the stenting procedure. To study this problem we have taken a multidisciplinary approach, with a team including representatives from interventional radiology, thoracic surgery, pulmonology, radiation oncology, and anesthesia. Our approach includes three-dimensional reconstruction of the airway from helical CT imaging to help determine the feasibility and potential benefit of the procedure, and bronchography with bronchoscopy at the time of stent placement to help define the lesion. We use metallic endobronchial stents placed under fluoroscopy as the primary therapy for palliation of large airway obstruction from lung cancer, without preceding laser ablation of the obstructing tumor mass. Presented at the Poster Session of the Thirty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Feb 3 5, Address reprint requests to Dr Zwischenberger, Division of Cardiothoracic Surgery, The University of Texas Medical Branch, 301 University Blvd, Galveston, TX ( jzwische@mspo2. med.utmb.edu). Material and Methods All patients on presentation had advanced lung cancer with severe dyspnea (American Thoracic Society dyspnea scale grade 4 [8]). Cases considered for stent placement were reviewed by a multiple disciplinary commit by The Society of Thoracic Surgeons /97/$17.00 Published by Elsevier Science Inc PII S (97)

2 1620 ZWISCHENBERGER ET AL Ann Thorac Surg STENT PLACEMENT FOR AIRWAY OBSTRUCTION 1997;64: Fig 1. Three-dimensional reconstruction of stenoses with proximal and distal airway measurements and length of stenotic region. tee consisting of representatives from thoracic surgery (J.B.Z.), interventional radiology (G.R.W., E.V.), and pulmonology (R.J.M.). Patients were selected based on the severity of their dyspnea and not on presumed outcome. Preprocedure staging confirmed the diagnoses of large airway obstruction in advanced unresectable lung cancer (stage IIIB or IV). The plan for adjunctive radiotherapy either before or after stent deployment was individualized by a radiation therapist dedicated to chest radiotherapy. All patients gave informed consent in accordance with institutional review board approval. Fourteen consecutive patients had a combination of fiberoptic bronchoscopy, chest roentgenography, CT scanning, helical CT with three-dimensional reconstruction, and intraluminal bronchography with selective bronchial catheterization under fluoroscopy to visually reconstruct and simulate the abnormal airway before and during stent placement. The three-dimensional image process included orienting the airway anatomy to measure the proximal and distal diameters of the airway in which the stent would be deployed and the length of the stenotic region. Three-dimensional images of the airways were created to assist with stent placement in the procedure room (Fig 1). At the time of the stent procedure, we used selective catheter techniques to perform bronchography of the obstructed airways (Fig 2). Curved catheters were placed over guidewires passed through the obstruction followed by injection of small amounts of contrast medium (Ethiodol; Savage Laboratories, Melville, NY) to outline the airways beyond the obstruction and the extent of the stenosis itself. We often placed guidewires into several adjacent bronchi to further outline the major bronchial branchpoints to determine how many stenoses there were, and the location and degree of these stenoses. The first 6 patients were stented in the operating room with a cardiothoracic anesthesiologist (S.K.A.) using a double-lumen endotracheal tube. The last 8 were stented in a radiology suite with general anesthesiologists and routine single-lumen tracheal intubation. Wallstent (Schneider [Europe] AG; Bülach, Switzerland) or Gianturco (Cook Inc, Bloomington, IN) intraluminal stents (Fig 3) were used alone or in combination (up to five stents) to establish patency of the distal trachea and the major bronchi. In selected cases with high-grade obstruction of the trachea, we used the safety guidewire technique by Fig 2. Bronchography of stenoses.

3 Ann Thorac Surg ZWISCHENBERGER ET AL 1997;64: STENT PLACEMENT FOR AIRWAY OBSTRUCTION 1621 Fig 3. Wallstent (top) and Gianturco stent (bottom). inserting two guidewires through the obstruction, using one as a working guidewire and the other as a safety guidewire, allowing insertion of a smaller tube to maintain access to the airway in case the endotracheal tube inadvertently slipped out. Upon completion of stent deployment, intraluminal patency was confirmed both by bronchography under fluoroscopy and by fiberoptic bronchoscopy. Likewise, bronchoscopic suctioning and bronchoalveolar lavage were performed before transfer from the procedure suite. After the procedure, the patients were extubated once awake and able to clear secretions. After stenting, daily chest roentgenograms assessed lung expansion (Fig 4), and CT scanning was used to assess patency of the airway (Fig 5). Chest physiotherapy, bronchodilator inhalation therapy, and early ambulation were part of the postprocedure protocol. Patients were discharged to hospital, hospice, or home care depending on their post-stenting functional status. Each outpatient was seen 2 weeks after discharge by the thoracic surgeon, then followed up for 6 months or until time of death. Functional status was assessed using the Karnofsky score. Dyspnea was assessed by the American Thoracic Society dyspnea scale. The patient s perception of maximum benefit was assessed using a questionnaire. If the patient was deceased, the date and cause of death were determined. The hospital charges and length-of-stay attributable to stent placement were obtained from hospital accounting. Results Fourteen patients underwent endobronchial stenting for palliative relief of severe dyspnea (American Thoracic Society grade 4) due to malignant large airway obstruction. The patients treated included 7 women and 7 men, with ages ranging from 38 to 76 years (mean, 58 years). The histology of the underlying malignancies included poorly differentiated non small cell (6), adenocarcinoma Fig 4. Sixty-four year old patient with bronchogenic carcinoma who presented with dyspnea. (A) Chest roentgenogram reveals a large mediastinal mass that surrounds and narrows the trachea to 5 mm in diameter. (B) A metallic stent (arrows) has been inserted that opens the trachea to 14 mm in diameter. The patient s dyspnea subsequently resolved. (2), squamous cell (4), large cell (1), and small cell (1). Clinical staging of the malignancy was IIIB in 10 patients and stage IV in 4 patients. Two patients also complained of mild hemoptysis. Before stenting, 9 patients had Karnofsky scores in the range of 50 to 70, with 2 greater than 70 and 3 less than 50. Seven patients received adjunctive radiation therapy (3 before and 4 after stenting). Indications for stent placement included severe dyspnea in all patients, plus bronchopleural fistula (1), tracheoesophageal fistula (1), and postoperative pneu-

4 1622 ZWISCHENBERGER ET AL Ann Thorac Surg STENT PLACEMENT FOR AIRWAY OBSTRUCTION 1997;64: monitis (4). The bronchi were completely obstructed in 4 patients before stenting. All patients had successful stent deployment with initial relief of airway stenosis ( 75% predicted diameter). Up to five stents were placed in a single patient. Stents were placed in the trachea (1 patient), main bronchi (4 right, 5 left, 1 bilateral), combination trachea and bilateral main bronchi (1), combination right main and right upper lobe bronchi (1), and right upper lobe bronchus (1). No procedural complications were noted. Technical challenges during stent placement included stent foreshortening with initial imprecision of placement, misinterpretation of bronchography (mucous versus tumor), and airway maintenance during stent manipulation. Upon successful stent deployment, vigorous pulmonary physiotherapy allowed resolution of atelectasis or infiltrate in 5 patients. No patient received additional stents after the initial procedure. Stent placement improved the dyspnea score in 7 patients (50%), with symptoms unchanged in the remainder. Upon follow-up questioning, at the time of perceived maximum benefit, 6 patients felt they had achieved significant improvement in functional status. Of the 9 patients surviving greater than 2 months, the Karnofsky score improved in 4 and was unchanged in 5. Ten of the patients were deceased at 8-month followup. Five patients died in less than 2 months. All 4 patients with stage IV disease died within 2 months of the procedure. Of those who died, 5 were able to return home before death, 4 never left the hospital or required early readmission, and 1 died at a nursing care facility. New problems identified in individuals before death included a tracheoesophageal fistula, ipsilateral pneumothorax, tracheostomy, and atrial fibrillation. Total length of stay ranged from 3 to 22 days (average, 10.2 days), and the average charge for the hospitalization was $22,067, excluding physician time for the three faculty in attendance during stent placement. Length of stay attributable to the procedure averaged four days. The cost of a single Wallstent is $1,000, whereas a Gianturco stent is $500 to 700. Fig 5. (A) Prestent computed tomographic scan at the level of the aortic arch showing a large mediastinal mass causing severe compromise of the airway with a high-grade stenosis (5-mm diameter) of the trachea. (B) Post-stent computed tomographic scan showing successful stent placement and patency of the trachea measuring 14 mm in diameter. Comment Lung cancer remains a major health problem in the Untied States, with more than 120,000 lung cancer deaths occurring annually. Although surgical resection remains the treatment of choice, many patients present at an advanced stage or have lung dysfunction which precludes an operation. Treatment with external beam radiotherapy, chemotherapy, or both is often performed with the hope of palliation of symptoms. When a major airway is occluded or narrowed due to malignancy, dyspnea may develop with respiratory failure, recurrent pneumonia, or atelectasis [9 11]. In such cases the goal is to achieve relief of the obstruction within days. Brachytherapy may be effective for small endobronchial lesions, but is ineffective for extrinsic compression or larger lesions. Laser photocoagulation and cryotherapy have emerged as potential treatment options for malignant airway obstruction; however, recurrence rates are high, and these modalities are ineffective for complete airway occlusion and for treating extrinsic compression [12, 13]. The limitation of these therapies for dealing with extrinsic airway compression has led to significant recent interest in the technique of endobronchial stenting [14]. With complete or near-complete malignant airway occlusion, CT scanning and bronchoscopy may give inadequate information to guide decision-making regarding the amount of viable lung distal to the obstruction or the technical feasibility of the stenting procedure. We established a protocol using helical CT scanning, bronchoscopy, and bronchography, with a multidisciplinary effort to evaluate patients and place metal endobronchial

5 Ann Thorac Surg ZWISCHENBERGER ET AL 1997;64: STENT PLACEMENT FOR AIRWAY OBSTRUCTION 1623 stents to palliate dyspnea caused by malignant large airway obstruction. Using this approach, we demonstrated successful airway stenting resulting in subjective short-term benefit to most patients, and improvement of dyspnea in half the patients treated. Although the idea for airway stenting can be traced back as far as the 1800s [15], only fairly recently has technology evolved to allow stenting to become a viable treatment option to relieve malignant airway obstruction. In 1952, Harkins [16] relieved a benign tracheal stenosis with a metal alloy tube. In the 1960s, Montgomery [17] used a siliconized T-shaped prosthesis in the trachea. The tracheal prosthesis was later modified into a tracheobronchial prosthesis by Westaby and associates [18]. The major problem was finding a material that would not lead to infection or foreign body reaction and could be placed with minimal trauma and complications. The ideal stent needs to fit securely in the airway lumen without requiring external fixation and without causing mucosal injury. Satisfactory expansion of the stricture must occur regardless of the origin or the forces causing extrinsic compression. The formation of granulation tissue secondary to airway inflammation should be minimal, and surrounding vascular and cartilaginous structures should not be damaged. Most importantly, the stent must be able to fit the entire stricture, be resistant to extrinsic compression, and allow manipulation in case of obstruction from tumor overgrowth or secretions [19]. Endobronchial stenting may be indicated to open large airways severely narrowed or obstructed due to a tumor mass when symptoms are significant and attributable to the obstruction, when the patient is not a candidate for surgical treatment, primary radiation, or chemotherapy, or when those therapies are not likely to be effective alone. Flexible fiberoptic bronchoscopy has been used extensively to diagnose, perform biopsy of, and stage tumors. Bronchoscopy allows examination of the airway and evaluation of the extent of tracheobronchial stenosis, and helps estimate distances from major branch points in the airway. Direct visualization also allows evaluation of extent, severity, and complexity of the stenosis in addition to obtaining specimens for further diagnosis [19]. Additional benefits of bronchoscopy include improved ventilation after secretion aspiration, control of bleeding, and, if necessary, endotracheal intubation. Unfortunately, the bronchoscopic image thoracic surgeons and pulmonologists are so familiar with can be misleading and may only show the actual point of obstruction or what we call the tip of the iceberg. Under fluoroscopic guidance, bronchoscopy gains a whole new level of understanding. The view through the fiberoptic bronchoscope is two-dimensional; however, when a bronchoscope is passed during fluoroscopy, the correlative image aids in identifying the exact anatomic location of the site under consideration. Helical CT with three-dimensional reconstruction has proved a very valuable addition in these patients, allowing us to plan stent positioning before the procedure. Helical CT also provides information on the extent of the tumor versus the extent of atelectatic lung. The benefit of three-dimensional reconstruction is limited when no distal airway is visualized. When this occurs, only the proximal airway diameter can be determined. After chest roentgenography, fiberoptic bronchoscopy, CT scanning, and helical CT with three-dimensional reconstruction we performed intraluminal bronchography (often with intraluminal guidewires to show bronchial branchpoints) to further image the airway. Intraluminal bronchography was especially useful to differentiate inspissated mucous from tumor. Once all of the imaging modalities were completed, the thoracic surgeon or pulmonologist and the interventional radiologist combined their perceptions of the airway anatomy to develop a best management plan for stent placement. This included stent location, how many stents to insert, and stent length. If crossing a branch takeoff of a bronchus, we prefer to insert a Gianturco stent. In the case of a bronchopleural or esophageal fistula within the airway, we found covered stents to be preferred; otherwise, the uncovered stents are technically easier. If two adjacent bronchi are to be stented, a double-barrel side-by-side approach deploying the stents simultaneously is used. In our experience and others, the two most commonly used metallic stents are the Gianturco Z stent and the Wallstent. The larger Gianturco stent [20] has small hooks that allow anchoring into the airway wall. However, the smaller Gianturco stents our group uses in the main or segmental bronchi do not have these small hooks. The Wallstent is a self-expandable stent, which expands and shortens after release. Easily adaptable, it is effective in treating pure extrinsic compression. The advantages and disadvantages of the Gianturco stent and the Wallstent have been reviewed previously [19, 21]. A multicenter study using the Wallstent for palliation of inoperable tracheobronchial cancer included 40 patients at seven centers. A total of 50 stents were inserted, of which 44 were determined to be adequately placed. The six suboptimally inserted stents were due to poor placement (n 2), immediate migration (n 2), and complications with anesthesia (n 2). The authors of that study concluded the Wallstent was the stent of choice for pure extrinsic stenoses and esophagobronchial fistulas. Although the rate of minor complications was shown to be high, the risk of migration was low when the diameter was correctly determined before insertion. Gianturco stent (n 35) and Wallstent (n 39) implantation also were compared by Rousseau and co-workers [22]. Eightynine percent of their patients showed symptomatic improvement after stent placement. Thirty-one percent of patients undergoing Gianturco stent placement experienced complications of migration, breaks in the filament, obstruction, or perforation. Complications involving the spontaneous fracture of Gianturco stents [23], vascular perforations, and granulation tissue formation [24] also have been reported. The main anesthesia considerations are to keep the patient adequately hydrated, monitor the hemodynamics and airway pressures, and maintain a patent airway throughout the procedure. Under general anesthesia, the degree of stenosis can increase due to either a loss of lung

6 1624 ZWISCHENBERGER ET AL Ann Thorac Surg STENT PLACEMENT FOR AIRWAY OBSTRUCTION 1997;64: volume or increased tumor compression of the airway without muscle tone. A double-lumen tube or long, small-diameter endotracheal tube should be immediately available during stent placement. Furthermore, there is a risk of significant bleeding in the airway [25], again necessitating a double-lumen tube to prevent aspiration into the uninvolved airway. Upon completion of the procedure, extubation has to be expedited so secretions can be coughed up immediately, as the stent may expose the remainder of the airway to inspissated secretions. We used a double-lumen endotracheal tube to allow uninterrupted ventilation of the unaffected lung and a working channel for the obstructed airway. Once we realized a double-lumen endotracheal tube was difficult to insert with large airway obstruction and further precluded tracheal access we began using a large-bore single-lumen tube. This greatly eased airway management and provided a large working channel for suctioning, guidewire placement (including multiple wires), and stent deployment while still allowing a sufficient lumen for ventilation. Insertion was to just below the cords. A large-bore single-lumen endotracheal tube is now our airway management technique of choice for stent insertion. In our experience, the Wallstent is easier to deploy because it is preloaded, whereas the Gianturco stent takes some manipulation to modify the delivery system for stent deployment. The Wallstent characteristically foreshortens, which can be difficult to estimate and occasionally requires placement of another stent. Although migration of the Giantuco stent has been noted, we have not seen migration with either stent. In those patients with treatable cancer, migration may occur after tumor shrinkage caused by radiation or chemotherapy. Another complication that has been described is breaking of struts or metal strands that hold the stents together. One patient presented to the clinic secondary to coughing up a piece of short wire. This complication had no further clinical consequences. A theoretical consideration is incompatibility of both Wallstents and Gianturco stents in the same patient. We have mixed stents for tailored use in the same patient with no complications. Another technical consideration we have seen is stents protruding back and hanging over the carina, for example, a left main bronchial stent hanging slightly over the right main bronchus. When this has happened, we have used a balloon to dilate the stent, thereby further opening the stent with the goal of slight antegrade movement to reduce or eliminate the overhang. Tumor flaps or a ball valve effect can cause air trapping with subsequent pneumothorax or hemodynamic instability. The question of where to perform these procedures and who should perform them has been asked [26]. Large, specialized centers, with the resources, experience, and personnel, are fully capable of performing this procedure. A multidisciplinary team of specialists (thoracic surgeon, interventional radiologist, anesthesiologist, pulmonologist, oncologist) should be used. Each member brings his or her own expertise, which results in the best care and outcome for the patient. Our group, using both the Gianturco stent and the Wallstent, believe both provide significant relief of obstruction with minimal complications. The question of which one to use is a decision best made by the multidisciplinary team after careful evaluation of the patient. The question of timing and role of radiation therapy was not addressed by our experience. We used radiation as best therapy to achieve palliation. To address this issue, a prospective, randomized study with and without radiation would be necessary. To date, we have successfully deployed stents and achieved airway patency in all patients in whom we have attempted the procedure. The techniques we describe allow accurate airway simulation, intraluminal stent placement, and relief of stenosis with an immediate risk that appears low. The major goal of this procedure was palliation of symptoms, and as such it was successful in achieving short-term perceived benefit in most patients. The early mortality of this patient population remains high, and it is not clear if stenting affects survival; however, mortality data presented here are similar to those of other studies of stenting malignant airway obstruction [21]. The patients with stage IV disease appeared to benefit least from stent placement. In the 5 patients who died in the first 2 months, the pre-stenting Karnofsky score was not predictive of outcome. Impact on functional status and dyspnea were modest for the group as a whole, yet dramatic in several cases. In addition to refining technical aspects of stent placement, future efforts should be directed toward determining which patients are most likely to benefit from airway stenting. We are currently undergoing clinical trials to define these patient populations prospectively. Ethical issues regarding the cost and potential benefit of palliative care at the end of life should be further addressed before the role of stenting for malignant airway obstruction may be properly established. References 1. Tojo T, Iioka S, Kitamura S, et al. Management of malignant tracheobronchial stenosis with metal stents and Dumon stents. Ann Thorac Surg 1996;61: Bolliger CT, Probst R, Tschopp K, Soler M, Perruchoud AP. Silicone stents in the management of inoperable tracheobronchial stenoses. Indications and limitations. Chest 1993; 104: Vergnon JM, Costes F, Bayon MC, Emonot A. Efficacy of tracheal and bronchial stent placement on respiratory functional tests. Chest 1995;107: Martinez-Ballarin JI, Diaz-Jimenez JP, Castro MJ, Moya JA. Silicone stents in the management of benign tracheobronchial stenoses. Tolerance and early results in 63 patients. Chest 1996;109: Sawada S, Tanigawa N, Kobayashi M, Furui S, Ohta Y. Malignant tracheobronchial obstructive lesions: treatment with Gianturco expandable metallic stents. Radiology 1993; 188: Sutedja G, Schramel F, van Kralingen K, Postmus PE. Stent placement is justifiable in end-stage patients with malignant airway tumours. Respiration 1995;62: Wilson GE, Walshaw MJ, Hind CR. Treatment of large airway obstruction in lung cancer using expandable metal

7 Ann Thorac Surg ZWISCHENBERGER ET AL 1997;64: STENT PLACEMENT FOR AIRWAY OBSTRUCTION 1625 stents inserted under direct vision via the fibreoptic bronchoscope. Thorax 1996;51: Brooks SM. Task group on surveillance for respiratory hazards in the occupied setting. Surveillance for respiratory hazards. ATS News 1982;8: George PJ, Rudd RM. Respiratory stents. Br J Hosp Med 1992;47: Hetzel MR, Smith SG. Endoscopic palliation of tracheobronchial malignancies. Thorax 1991;46: Mostovych M, Mathisen DJ. Management of malignant airway obstruction. In: Pass HI, Mitchell JB, Johnson DH, Turrisi AT, eds. Lung cancer: principles and practice. Philadelphia: Lippincott-Raven, 1996: Cavaliere S, Foccoli P, Farina PL. Nd:YAG laser bronchoscopy. A five-year experience with 1,396 applications in 1,000 patients. Chest 1988;94: Walsh DA, Maiwand MO, Nath AR, Lockwood P, Lloyd MH, Saab M. Bronchoscopic cryotherapy for advanced bronchial carcinoma. Thorax 1990;45: Mathisen DJ, Grillo HC. Endoscopic relief of malignant airway obstruction. Ann Thorac Surg 1989;48: Trendelenburg F. Beitraege zu den Operationen an den Luftwegen. Langenbecks Arch Chir 1872;13: Harkins WB. An endotracheal metallic prosthesis in the treatment of stenosis of the trachea. Ann Otol Rhinol Laryngol 1952;61: Montgomery WW. T-tube stent. Arch Otolaryngol 1965;83: Westaby S, Jackson JW, Pearson FG. A bifurcated silicone rubber stent for relief of tracheobronchial obstruction. J Thorac Cardiovasc Surg 1982;83: Colt HG, Dumon JF. Airway stents. Present and future. Clin Chest Med 1995;16: Uchida BT, Putnam JS, Rosch J. Modifications of Gianturco expandable wire stents. AJR 1988;150: Monnier P, Mudry A, Stanzel F, et al. The use of the covered Wallstent for the palliative treatment of inoperable tracheobronchial cancers. A prospective, multicenter study. Chest 1996;110: Rousseau H, Dahan M, Lauque D, et al. Self-expandable prostheses in the tracheobronchial tree. Radiology 1993;188: Schäfers H-J, Hamm M, Wagner TOF. Gianturco selfexpanding metallic stents [Letter]. Eur J Cardiothorac Surg 1992;6: Rauber K, Franke C, Rau WS. Self-expanding stainless steel endotracheal stents: an animal study. Cardiovasc Intervent Radiol 1989;12: Fujii K, Okida M, Fujioka Y, Kobayashi M, Fujimoto M. General anesthesia for expandable endotracheobronchial metal stent insertion. Jpn J Anesthesiol 1996;45: Coolen D, Slabbynck H, Galdermans D, Van Schaardenburg C, Mortelmans LL. Insertion of a self-expandable endotracheal metal stent using topical anaesthesia and a fibreoptic bronchoscope: a comfortable way to offer palliation. Thorax 1994;49:87 8.