Metallic Stent Placement for the Management of Tracheal Carina Strictures and Fistulas: Technical and Clinical Outcomes

Transcription

1 Vascular and Interventional Radiology Original Research Kim et al. Outcomes of Tracheal Stent Placement Vascular and Interventional Radiology Original Research Jinoo Kim 1 Ji Hoon Shin 2 Jin-Hyoung Kim 2 Ho-Young Song 2 Soon-Young Song 3 Choong Ki Park 4 Kim J, Shin JH, Kim JH, Song HY, Song SY, Park CK Keywords: airway, carina, fistula, stent, stricture DOI: /AJR Received December 7, 2012; accepted after revision June 27, Department of Radiology, Ajou University School of Medicine, 164 Worldcup-ro, Yeongtong-gu, Suwon , Korea. 2 Department of Radiology and Research Institute of Radiology, University of Ulsan, College of Medicine, Asan Medical Center, 86 Asanbyeongwon-gil, Songpa-gu, Seoul , Korea. Address correspondence to J. H. Shin (jhshin@amc.seoul.kr). 3 Department of Radiology, Hanyang University, College of Medicine, Hanyang University Seoul Hospital, Seoul, Korea. 4 Department of Radiology, Hanyang University, College of Medicine, Hanyang University Guri Hospital, Gyeonggi-do, Korea. AJR 2014; 202: X/14/ American Roentgen Ray Society Metallic Stent Placement for the Management of Tracheal Carina Strictures and Fistulas: Technical and Clinical Outcomes OBJECTIVE. The purpose of this article is to assess the technical and clinical outcomes of metallic stent placement in strictures and fistulas involving the carina. MATERIAL AND METHODS. We performed a retrospective analysis of patients who had undergone stenting for disease involving the carina. We initially reviewed the symptoms, underlying causes, and the types of stent configuration used. We also assessed the technical success rate of stenting, its effectiveness in achieving symptomatic relief, the incidence of stent-related complications, and stent patency. RESULTS. Thirty-two stenting procedures were performed in 23 patients (mean age, 56.3 years) for the treatment of strictures (n = 21), an esophagorespiratory fistula (n = 1), or both (n = 1) present in the carina. Three cases were associated with benign causes, whereas 20 were related to malignancies. Dyspnea was the most common symptom (n = 22). We placed metallic stents in four different configurations, among which placement in juxtacarinal segments was the most common configuration (n = 23). Technical success was achieved in 96.9% of cases, and symptomatic improvement was observed in 90.6% of cases. Stent-related complications were observed after 10 procedures (31.3%). Stent obstruction occurred in seven patients (21.9% of procedures), most commonly because of tumor progression. The mean follow-up period was 83.1 days, during which time 15 patients died as a result of disease progression, five were discharged without hope for improvement, two were discharged without symptomatic recurrence, and one was lost to follow-up. CONCLUSION. Airway stenting can be performed in the carina with high technical success using variable stent configurations. Although the rate of immediate symptomatic improvement is high, stent-related complications frequently occur. A irway stenting has been shown to provide prompt relief of airway obstruction caused by both benign and malignant strictures [1 6]. However, because airway stenting is associated with a high rate of complications, such as stent migration, obstruction, and sputum retention, the procedure is most often reserved for palliative management of malignant strictures in patients with otherwise-exhausted treatment options and short life expectancies [7 11]. Currently, there are accumulating data in the published literature regarding the outcome of airway stenting, and various stent designs (e.g., silicone and bare or covered self-expandable metallic stents) and techniques for their placement (e.g., bronchoscopic or fluoroscopic guidance and permanent or temporary stenting) have been pro- posed to overcome the challenges that are frequently faced in airway stenting [8, 12, 13]. Regarding disease involving the carina, the anatomy of the airway in this region, as well as the location or pattern of the lesions involving it, add to the complexity of stent placement [14]. Both benign and malignant causes may result in the development of strictures or esophagorespiratory fistulas (ERFs) in the carina that do not respond to other modes of therapy. When stent placement is the only available option, careful preprocedural planning with respect to the stent selection and its location in the airway is as important as the technical aspects of the procedure itself, to achieve sufficient coverage of the lesion without compromising patent airways. In this study, we evaluated the outcome of airway stenting for 880 AJR:202, April 2014

2 Outcomes of Tracheal Stent Placement strictures and ERF involving the carina using various stent configurations. Materials and Methods Patients Our institutional review board approved this retrospective study, and the requirement for informed consent was waived. From June 1998 to December 2011, a total of 243 patients underwent placement of airway stents under fluoroscopic guidance. The patients we selected from this group were those who had airway stents placed because of benign or malignant disease involving the carina. As a result, 23 patients (14 men and nine women) ranging in age from 23 to 76 years (mean age, 56.3 years) composed the final study population analyzed in this retrospective study. The clinical indications for stenting and the underlying disease of each patient were retrospectively assessed, together with the findings determined on bronchoscopy and radiologic studies. Stent Placement The stenting procedure was performed by three board-certified interventional radiologists with 6 26 years of clinical experience in airway intervention. Routine premedication included IV administration of sedatives and topical anesthesia of the pharynx and larynx using lidocaine aerosol spray. After a patient was positioned in the supine position on the fluoroscopy unit, primary assessment of the airway was performed by a pulmonologist using a bronchoscope that was kept on standby in the angiography suite. After identifying a lesion, a inch, hydrophilic guidewire (Radifocus, Terumo) was passed through the bronchoscope and positioned distal to the lesion. The bronchoscope was then removed, leaving the guidewire in place. Under fluoroscopic guidance, a sizing catheter with multiple side holes (Royal Flush II angiographic catheter, Cook Medical) was then passed over the guidewire and positioned across the lesion, after which a small amount (usually < 5 ml) of water-soluble iodinated contrast medium (mixed with lidocaine at a ratio of 1:1 to prevent bronchospasm) was hand-injected through the catheter to outline the location and pattern of disease around the carina. The stent was then delivered into the airway over the prepositioned guidewire. The stents were placed in four different configurations, which are presented in Figure 1. Immediately after stent placement, bronchoscopy was performed on site to confirm that the stents had been appropriately positioned. Fig. 1 Four different stent configurations placed in carina. Type A, juxtacarinal, involves one to three selfexpandable metallic stents placed in juxtacarinal segments of trachea or in either or both main bronchi. Type B, hinged, involves metallic stents placed in lower trachea and either of main bronchi and held together by hinge. With type C, flared, distally flared metallic stent with flared portion is placed deep into carina. Type D, unilateral, is straight tracheobronchial metallic stent extending from lower trachea unilaterally into either of main bronchi. Outcome and Follow-Up Technical success was defined as successful placement of the stent at the planned location that was verified on site by fluoroscopy and bronchoscopy. Clinically, we assessed the rate of symptomatic improvement after the procedure by comparing the symptoms for which stenting was indicated before and after the procedure. Routine follow-up, as a means to assess possible stent migration or underexpansion of the stent, included chest radiography performed within 3 days of the procedure and both fluoroscopy and bronchoscopy performed within 1 week after the procedure. Any of these follow-up studies was performed irrespective of the planned schedule if warranted by the development of adverse clinical conditions. We also analyzed the rate of stent-related complications, patency of the stents, and the final prognosis for each patient. Results Patients Among the 23 patients who underwent airway stenting, 21 (91.3%) had strictures, one (4.3%) had an ERF, and one other (4.3%) had both (Table 1). The underlying causes in- TABLE 1: Demographic Data of 23 Patients Who Underwent Airway Stenting for the Management of Lesions in the Carina Characteristic No. (%) of Patients (n = 23) Type of lesion Stricture 21 (91.3) ERF 1 (4.3) Both stricture and ERF 1 (4.3) Cause Benign 3 (13.0) Postintubation (stricture) 1 (4.3) Postsurgical (stricture) 1 (4.3) Postradiation (fistula) 1 (4.3) Malignant 20 (87.0) Lung cancer (stricture) 15 (65.2) Esophageal cancer (fistula and stricture) 2 (8.7) Metastatic lymphadenopathy (stricture) 3 (13.0) Clinical presentation Dyspnea 22 (95.7) Cough 7 (30.4) Obstructive pneumonia 5 (21.7) Hemoptysis 1 (4.3) Note ERF = esophagorespiratory fistula. AJR:202, April

3 Kim et al. cluded both benign (n = 3; 13.0%) and malignant (n = 20; 87.0%) diseases. The details are presented in Table 1. The most common clinical symptom was dyspnea (n = 22; 95.7%) followed by others such as cough (n = 7; 30.4%), obstructive pneumonia (n = 5; 21.7%), and hemoptysis (n = 1; 4.3%) (Table 1). Two or more symptoms coexisted in 11 patients. Six patients had histories of unsuccessfully attempted bronchoscopic intervention caused by failed passage of the endoscope through the tight strictures in three patients, recurrent strictures after balloon dilation or laser ablation in two patients, and a stricture that was unresponsive to balloon dilation in one patient. Stent Placement We performed a total of 32 stenting procedures for 23 patients; two procedures were performed for five patients and three were performed for two patients because of either symptom recurrence or the development of stent-related complications. The most commonly used stent configuration was type A (n = 23; 72%) (Fig. 2), followed by type B (n = 4; 13%), type D (n = 3; 9%), and type C (n = 2; 6%) (Table 2). Regardless of the stent configuration, the stents were covered metallic stents (with silicone covering in two cases and polytetrafluoroethylene covering in the remainder) in all but one case of type A configuration, in which bare metallic stents were placed in the juxtacarinal trachea and the left main bronchus to manage a malignant stricture. Technical success of stent placement in the carina was achieved in 31 cases (96.9%) (Table 3). The single case of technical failure was observed in a patient who underwent type A configuration stenting because of a malignant stricture in the carina and where the stent in the left main bronchus was malpositioned during placement; the malpositioned stent was immediately removed on site and the procedure was discontinued because of concerns that the patient s condition could worsen during a lengthy procedure. Outcome and Follow-Up Symptomatic improvement was observed after 29 procedures (90.6%) (Table 3). Excluding the aforementioned case of technical failure, symptomatic improvement could not be achieved in two patients. Because one of these patients had severe pneumonia at the time of the procedure, her respiratory symptoms were not resolved with stent placement; this patient eventually died of sepsis. The other patient had undergone two stenting procedures, the second of which was performed because of migration of the first stent. The patient was in an altered mental status after being previously resuscitated from respiratory arrest, and, for unidentifiable reasons, the second stent failed to relieve his respiratory distress. Of the two patients with ERF, a radiationinduced fistula was successfully sealed in one patient by a stent that was temporarily Fig year-old woman with malignant stricture caused by local extension of esophageal cancer. A and B, CT performed before stent placement shows esophageal cancer invading lower trachea and bilateral main bronchi (A). There is severe stricture in carina and left main bronchus (B). C and D, Two covered metallic stents (type A, juxtacarinal configuration) were placed in juxtacarinal airway, one in lower trachea (C) and other in left main bronchus (D), resulting in immediate relief of dyspnea and gradual improvement of pneumonia in left lobe. E and F, Follow-up CT (E) shows patent trachea and left main bronchus after stenting, whereas progressive tumor extension into unstented right main bronchus was seen. Note esophageal stent that had subsequently been placed (F). placed for less than 2 weeks, whereas a fistula associated with esophageal cancer in the other patient was unsuccessfully managed owing to migration of the stent 2 hours after placement (Fig. 3); this stent was consequently removed. Both patients with benign strictures showed symptomatic improvement after stenting. Although the stent was successfully removed after 4 weeks in one of these patients, the stent in the other patient TABLE 2: Stent-Related Complications in Four Different Stent Configurations Placed in the Carina Complication Type A (Juxtacarinal) Type B (Hinged) Type C (Flared) Type D (Unilateral) Total (n = 32), no. (%) 23 (72) 4 (13) 2 (6) 3 (9) Complications (n = 10 [31.3%]) Migration (n = 4) Sputum retention (n = 4) Hemoptysis (n = 2) Stent obstruction (n = 7 [21.9%]) Tumor progression (n = 5) Granulation tissue (n = 1) Impacted sputum (n = 1) Note Except where noted otherwise, data are number of stents. 882 AJR:202, April 2014

4 Outcomes of Tracheal Stent Placement could not be retrieved because of granulation tissue formation, which was observed 18 weeks after stent placement. Stent-related complications occurred after 10 procedures, thus yielding a complication rate of 31.3% (Table 2). The complications were stent migration (n = 4), sputum retention (n = 4), and minimal transient hemoptysis (n = 2). The four cases of stent migration were observed in type A (n = 3) and type B (n = 1) configurations. The migrated stents were removed in three cases followed by subsequent insertion of new stents, whereas in the fourth case, an additional stent was superimposed on the migrated stent without the latter being removed. With regard to stent patency (Table 2), obstruction of the stent lumen was seen in seven patients (21.9% of procedures) and was attributed to the progression of underlying malignancies (n = 5), benign granulation tissue (n = 1), and impacted sputum (n = 1). Malignant obstruction of the stent was observed only in the type A configuration and included a single case of bare metallic stent placement. With the covered stents, luminal obstruction due to stent ingrowth was observed at either or both ends of the stent, whereas with the bare stent, tumor ingrowth occurred through the cells between the stent struts. The follow-up period ranged from 3 to 357 days (mean, 83.1 days), during which 15 (65.2%) patients died as a result of progression of underlying diseases; five patients (21.7%) were discharged without hope for improvement, two (8.7%) were discharged without recurrence of symptoms, and one (4.3%) was lost to follow-up (Table 3). Discussion Ideally, when stents are placed in the carina, they should sufficiently cover the entire lesion, whether it is a stricture or fistula, without compromising the patency of the trachea or of the main bronchi. However, realistically, because the procedure is more challenging in such an anatomically complex region, a trade-off between total lesion coverage and preservation of airway patency may be inevitable. Therefore, in our study, incomplete coverage of malignant strictures by the stents was noted in patients with extensive lesions spreading distally beyond the carina. This may explain the high rate of stent obstruction caused by tumor progression, as seen at follow-up. Nonetheless, when immediate relief of obstructive symptoms is the primary objective, which is usually the case in patients referred for airway stenting, stent placement may be the best and only available palliative option, even if full coverage of the lesion cannot be achieved. To overcome the anatomic difficulties in the carina, some stent designs for this anatomic location have been described in the literature, such as the Dumon Y-stent, silicone Y-stent, and the double Y-stent, among others [14 17]. In contrast to these types of stents, the stents used in our study were all self-expandable metallic stents, which are advantageous in that they are easy to deliver, conform well to the anatomy of the native airway lumen, and are associated with a lower rate of migration [1, 12]. To the best of our knowledge, there is currently no large study focused exclusively on stent placement in the carina. Various metallic stent configurations were used in our study patient group, and we categorized them into four different types. Type A (juxtacarinal) and type B (hinged) stent configurations were identical to each other except that the stents in the trachea and bronchus were TABLE 3: Outcome of Stent Placement in the Carina Outcome No. (%) of Procedures or Patients Procedure outcome (n = 32) Technical success 31 (96.9) Symptomatic improvement 29 (90.6) Patient outcome (n = 23) Death 15 (65.2) Discharged without hope for 5 (21.7) improvement Discharged without recurrence 2 (8.7) Lost to follow-up 1 (4.3) Note The mean follow-up period was 83.1 days. Fig year-old man with history of surgery and concurrent chemoradiation therapy for esophageal cancer who developed malignant stricture and esophagorespiratory fistula in carina. A C, After verifying location of lesions with bronchoscopy, wire was passed through endoscope (A), over which stent (type B, hinged configuration) was placed (B and C). D, Esophagography performed after stent placement showed that fistula (arrow) had been sealed off by covered stent. However, chest radiograph performed 2 hours later because of sudden development of respiratory distress revealed distal migration of stent. E and F, Stent was subsequently removed using hookwire (E) and was followed by placement of endotracheal tube (F). Patient eventually developed worsening pneumonia in left lung and received ventilator care until he was discharged 42 days later without hope for improvement. AJR:202, April

5 Kim et al. placed separately around the carina in the former and attached to one another by a hinge in the latter. The hinged stent allowed the trachea and one side of the main bronchus to be stented during a single passage over the wire, thereby rendering the procedure more efficient and more tolerable for the patient. Despite our presumption that hinged stents would be associated with a lower risk of migration compared with separately placed stents, the stent was found to migrate in one of the four patients who underwent placement of a hinged stent. Although the carina itself was exposed (i.e., not covered by the stent) in both the type A and B configurations, the stents were designed so that the radial force on the margins of the stents in the trachea and the bronchus would improve the luminal patency in the stricture at the carina. The potential formation of stentinduced granulation tissue or the ingrowth of underlying malignancies in the exposed carina would be the major disadvantages of using these stents. The third type of stent, type C (flared), had a wide distal margin that could be placed deeper into the carina compared with the previous two configurations and without compromising either of the main bronchi. As a result, all but the lower edge of the carina would be covered by the stent. However, an insufficient safety margin or insufficient coverage of tumors extending distally into the main bronchus are disadvantages of such a stent configuration. In the type D (unilateral) configuration, the covered metallic stent in the distal trachea was placed across the carina into one side of the main bronchus. Because the orifice of the contralateral bronchus becomes totally covered by the stent, this procedure was performed only when the contralateral bronchus was already completely obstructed by the tumor and was associated with total lung collapse. Furthermore, these patients had respiratory symptoms attributable to the progressive stricture on the ipsilateral side and not associated with the atelectatic lung on the contralateral side. The major advantage of this configuration was that it offered total coverage of both the carina and the juxtacarinal airway segments, thereby potentially lowering the risk of tumor ingrowth at the stent margin. With the exception of the type D configuration, because there were no specific indications for the use of any of the configurations, the choice was made by the interventional radiologist performing the procedure. Regardless of the stent design, our study shows that placing stents in the carinas is technically feasible and can be performed with a high technical success rate. In all but one case in which the stent was malpositioned during its placement, the interventional radiologist had complete control of the stents during their positioning and deployment. Such meticulous technique was possible because of the guidance of fluoroscopy, through which the radiologists could gain a good understanding of the airway anatomy and accurately visualize the position of the radiopaque stents. In our case, cooperation with the pulmonologists who performed the bronchoscopy immediately before and after stent placement further enhanced the accuracy of stent positioning and allowed quick decisions to be made in the event of any complication occurring during the procedure. When the stent was either malpositioned or migrated during the procedure, the stent could be removed without difficulty under fluoroscopic guidance using a technique described elsewhere in the published literature [18, 19]. Temporary stent placement in the airway has been shown to be effective in managing ERF [13, 20 22]. Even though our study included only one case of a technically successful stenting procedure for the management of an ERF, the successful outcome with complete sealing of the fistula in this patient shows that temporary stent placement may be an effective treatment option in the management of fistulas occurring in the carina. Temporary stenting, which involves the removal of the stent after treatment, has also been shown to be effective for treating benign airway strictures. A short duration of stenting is associated with a lower risk of stent-related complications and allows the stent to be easily retrieved. Kim et al. [8] previously proposed the optimal period of temporary stenting for benign airway disease to be 6 months. In the present study, the stent could not be removed in one of the three patients with benign airway disease because of the development of granulation tissue at the edge of the stent, which was observed on endoscopy 18 weeks after its placement. In the two other patients, the stents were successfully removed within 1 month of their placement. With respect to the clinical outcome of stenting for malignant strictures in the carina, we focused on immediate symptomatic improvement rather than on the long-term outcomes and survival. However, most patients treated in this study were at the terminal stages of their malignancies and with short life expectancies, where immediate management of airway obstruction was the major objective rather than prolonging their survival. In this regard, a high rate of symptomatic improvement was achieved in our study group. Regarding the two cases in which immediate symptomatic relief was not achieved despite our technical success, the importance of selecting appropriate patients for airway stenting can never be overemphasized. Although we were not able to verify the cause of the stenting failure in our patients, they presumably had respiratory distress attributed to causes unrelated to their main airway obstruction (i.e., lobar pneumonia in one patient and respiratory depression secondary to altered mental status in the other). Despite our high technical success rate and the successful immediate outcome of stenting, stent-related complications and obstruction of the stent lumen were not rare in our group. Stent obstruction and migration and sputum retention are issues that have been addressed in the literature since the introduction of airway stenting [7, 9, 10]. Sputum retention, which occurred in four of our patients, is a result of diminished mucociliary clearance and is known to occur at a significantly higher rate when using covered stents rather than bare metallic stents [10, 19]. With regard to stent obstruction, partial exposure of the carina in type A, B, and C stent configurations and insufficient coverage of the lesions in any of the four configurations due to lesions extending into the distal main or segmental bronchi were possibly attributable to the high rate of tumor ingrowth. A high rate of stent migration was observed in our study patients, most occurring with the type A configuration. Such a distribution may be explained by the higher frequency of the type A configuration used compared with the other configurations (72% vs 28%). Furthermore, the numerous stents used in the type A configuration (i.e., between two and three in one patient) may also be the cause of the increased stent migration rate compared with that seen in single stent placement. There are a few limitations to our study. Its retrospective nature is the first limitation, because all of the data were reviewed on an electronic database and the strategic aspects of stent placement from the planning stage to follow-up were not part of a common protocol. There was also a lack of objective data, such as the results of pulmonary function tests. We think that this was because of the critical condition of the patients, most of whom were physically unable to undergo such studies. Another limitation is the lack of randomization regarding the type of stent 884 AJR:202, April 2014

6 Outcomes of Tracheal Stent Placement configuration that was selected. Instead, the stents were chosen according to the decision made by the interventional radiologist at the time of the procedure. Furthermore, a comparison of the outcomes among the different stent configurations was not possible because of the unevenly distributed frequency of procedures performed in each type of stent configuration, where the type A configuration constituted most of the procedures. In conclusion, stent placement for the management of benign and malignant lesions involving the carina is technically feasible and effective for achieving immediate symptomatic improvement. However, because stentrelated complications remain problematic, further investigations should be undertaken in an effort to decrease the complication rate. References 1. Wood DE, Liu YH, Vallieres E, Karmy-Jones R, Mulligan MS. Airway stenting for malignant and benign tracheobronchial stenosis. Ann Thorac Surg 2003; 76: Carré P, Rousseau H, Lombart L, et al. Balloon dilatation and self-expanding metal Wallstent insertion for management of bronchostenosis following lung transplantation. The Toulouse Lung Transplantation Group. Chest 1994; 105: Takamori S, Fujita H, Hayashi A, et al. Expandable metallic stents for tracheobronchial stenoses in esophageal cancer. Ann Thorac Surg 1996; 62: Saad CP, Murthy S, Krizmanich G, Mehta AC. Self-expandable metallic airway stents and flexible bronchoscopy: long-term outcomes analysis. Chest 2003; 124: Razi SS, Lebovics RS, Schwartz G, et al. Timely airway stenting improves survival in patients with malignant central airway obstruction. Ann Thorac Surg 2010; 90: Vonk-Noordegraaf A, Postmus PE, Sutedja TG. Tracheobronchial stenting in the terminal care of cancer patients with central airways obstruction. Chest 2001; 120: Stephens KE Jr, Wood DE. Bronchoscopic management of central airway obstruction. J Thorac Cardiovasc Surg 2000; 119: Kim JH, Shin JH, Song HY, Shim TS, Yoon CJ, Ko GY. Benign tracheobronchial strictures: longterm results and factors affecting airway patency after temporary stent placement. AJR 2007; 188: Gaissert HA, Grillo HC, Wright CD, Donahue DM, Wain JC, Mathisen DJ. Complication of benign tracheobronchial strictures by self-expanding metal stents. J Thorac Cardiovasc Surg 2003; 126: Zakaluzny SA, Lane JD, Mair EA. Complications of tracheobronchial airway stents. Otolaryngol Head Neck Surg 2003; 128: Vinograd I, Keidar S, Weinberg M, Silbiger A. Treatment of airway obstruction by metallic stents in infants and children. J Thorac Cardiovasc Surg 2005; 130: Chung FT, Chen HC, Chou CL, et al. An outcome analysis of self-expandable metallic stents in central airway obstruction: a cohort study. J Cardiothorac Surg 2011; 6: Witt C, Dinges S, Schmidt B, Ewert R, Budach V, Baumann G. Temporary tracheobronchial stenting in malignant stenoses. Eur J Cancer 1997; 33: Sonett JR, Keenan RJ, Ferson PF, Griffith BP, Landreneau RJ. Endobronchial management of benign, malignant, and lung transplantation airway stenoses. Ann Thorac Surg 1995; 59: Oki M, Saka H, Kitagawa C, Kogure Y. Silicone Y-stent placement on the carina between bronchus to the right upper lobe and bronchus intermedius. Ann Thorac Surg 2009; 87: Oki M, Saka H, Kitagawa C, et al. Double Y-stent placement for tracheobronchial stenosis. Respiration 2010; 79: Dutau H, Toutblanc B, Lamb C, Seijo L. Use of the Dumon Y-stent in the management of malignant disease involving the carina: a retrospective review of 86 patients. Chest 2004; 126: Shin JH, Song HY, Ko GY, et al. Treatment of tracheobronchial obstruction with a polytetrafluoroethylene-covered retrievable expandable nitinol stent. J Vasc Interv Radiol 2006; 17: Kim JH, Shin JH, Song HY, Lee SC, Kim KR, Park JH. Use of a retrievable metallic stent internally coated with silicone to treat airway obstruction. J Vasc Interv Radiol 2008; 19: Freitag L, Tekolf E, Steveling H, Donovan TJ, Stamatis G. Management of malignant esophagotracheal fistulas with airway stenting and double stenting. Chest 1996; 110: Kim JH, Shin JH, Song HY, Choi CM, Shim TS. Esophagorespiratory fistula without stricture: palliative treatment with a barbed covered metallic stent in the central airway. J Vasc Interv Radiol 2011; 22: Shin JH, Kim JH, Song HY. Interventional management of esophagorespiratory fistula. Korean J Radiol 2010; 11: AJR:202, April