- Email: [email protected]
Double stenting for esophageal and tracheobronchial stenoses
Double Stenting for Esophageal and Tracheobronchial Stenoses Hiroaki Nomori, MD, Hirotoshi Horio, MD, Yoshihiro Imazu, MD, and Keiichi Suemasu, MD Department of Thoracic Surgery, Saiseikai Central Hospital, Tokyo, Japan
Background. We examined the complications and outcomes of placing stents for both esophageal and tracheobronchial stenoses. Methods. We placed stents for both esophageal and tracheobronchial stenoses in 8 patients (7 with esophageal cancer and 1 with lung cancer). Covered or noncovered metallic stents were used for the esophageal stenoses, except in 1 patient treated with a silicone stent. Silicone stents were used for the tracheobronchial stenoses. The grades of esophageal and tracheobronchial stenoses were scored. Results. All patients experienced improvement of grades of both dysphagia and respiratory symptoms after stent therapy. The complications were: (1) 2 patients suffered respiratory distress after placement of the esophageal stent because of compression of the trachea by the stent; and (2) 3 patients developed new esophago-
tracheobronchial fistulae, and 2 patients had recurring fistula symptoms because of growth of preexisting fistulae after the stent placement, which were caused by pressure from the 2 stents. Despite the fistulae, the 5 patients treated with covered metallic stents did not complain of fistula symptoms, but 2 patients treated with noncovered metallic or silicone stents did complain. Conclusions. For patients with both esophageal and tracheobronchial stenoses, a stent should be introduced into the tracheobronchus first. Because placement of stents in both the esophagus and tracheobronchus has a high risk of enlargement of the fistula, a covered metallic stent is preferable for esophageal cancer involving the tracheobronchus.
R
esophageal cancer in 7 patients and lung cancer in 1. The mean length of esophageal stenoses in the esophageal cancer patients was 5.7 cm (range: 3–9). The sites of airway stenoses were the trachea in 5 patients, the left main bronchus in 2, and the right main bronchus in 1. All patients underwent chemoradiotherapy before the stent placements. In addition, one patient (patient 1) underwent also radiotherapy after the stent placement. Four patients (patient 5, 6, 7 and 8) already had fistulae before stent therapy. For tracheobronchial stenosis, we used silicone stents but not metallic stents to prevent a recurrence of stenosis due to tumor growth between the wires of the metallic stents [2]. Six Dumon stents (Novatech Co, Aubagne, France) were placed in 6 patients and 2 dynamic stents (Willy Ru¨sch AG, Kernen, Germany) in 2. Dumon stents were placed via an endotracheal tube under general anesthesia as described previously [4]. The dynamic stents were placed using specialized forceps under general anesthesia. The esophageal metallic stents were placed using the regular method under local anesthesia [5–7]. For esophageal stenosis, we generally used metallic stents but not silicone stents, because silicone stents tend to have a higher risk of esophageal injury and are also easily packed with food. Nine covered selfexpandable metallic stents (Covered Ultraflex Esophageal Stent System, Microinvasive, Boston Scientific Co, Boston, MA) were placed in 6 patients, 2 noncovered Ultraflex metallic stents in 2 patients, and 1 silicone stent
ecent advances in stent therapy have increased the quality of life of patients with inoperable cancer. Tracheobronchial stenoses have been treated by expandable metallic stent [1, 2] and silicone stent such as Dumon or dynamic stents [3–5]. Esophageal stenoses have classically been treated with plastic or silicone stent, but self-expandable metallic stent have recently been introduced for this purpose [5, 6]. Patients with advanced esophageal cancer often have respiratory distress as a result of tracheobronchial invasion, and this necessitates stent therapy not only for esophageal stenosis but also for tracheobronchial stenosis. In this report we examine the outcomes of stent therapy for both esophageal and tracheobronchial stenoses in 8 patients and recommend some management techniques that can be used to prevent complications.
Material and Methods From 1991 to 1999, we performed stent therapy for 28 patients with esophageal stenoses and 45 patients with tracheobronchial stenoses. Of these, 8 patients received stent therapy for both esophageal and tracheobronchial stenoses (Table 1). The causes of the stenoses were Accepted for publication May 14, 2000. Address reprint requests to Dr Nomori, Department of Surgery, Saiseikai Central Hospital, 1-4-17 Mita, Minato-ku, Tokyo 108-0073, Japan; e-mail: [email protected].
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2000;70:1803–7) © 2000 by The Society of Thoracic Surgeons
0003-4975/00/$20.00 PII S0003-4975(00)02042-7
1804
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
Ann Thorac Surg 2000;70:1803–7
Table 1. Patients Characteristics and Treatments for Esophageal and Airway Stenosis Airway Stent Patient No.
Age/ Gender
Cause of Stenoses
Site of Airway Stenoses
Fistula Before Stenting
Type
Time (months)a
Esophageal Stent Type (Numbers)
1 2
56/F 67/M
Es. Ca Es. Ca
Trachea Trachea
⫺ ⫺
Dynamic Dynamic
3 14
3 4 5
65/M 65/M 45/M
Es. Ca Es. Ca Es. Ca
Main Br. Trachea Trachea
⫺ ⫺ ⫹
Dumon Dumon Dumon
5 20 12
6 7
42/F 65/M
Es. Ca Es. Ca
Main Br. Main Br.
⫹ ⫹
Dumon Dumon
2 4
Covered MS (1) Covered MS (2)
8
76/M
Lung Ca
Trachea
⫹
Dumon
12
Covered MS (1)
a
Noncovered MS (1) Silicone stent (1) Noncovered MS (1) Covered MS (1) Covered MS (1) Covered MS (2)
Time (months)a 3 0 5 0 18 9 12 2 4 6 12
Time means a period between the diagnosis of the diseases and the stent placement.
Br ⫽ bronchus;
Ca ⫽ cancer;
Es ⫽ esophageal;
F ⫽ female;
(Flexible Artificial Esophagus, Sumitomo Bakelite Co, Ltd, Tokyo, Japan) in 1. The main purposes for using these esophageal stents were to dilate the stenoses in 6 patients (patients 1, 2, 3, 4, 6 and 8) and to block the esophagotracheobronchial fistula in the 2 (patients 5 and 7). A noncovered metallic stent or a silicone stent had to be used in 2 patients (patients 1 and 2) because a covered metallic stent was not commercially available in Japan at the time of treatment. In 5 patients (patients 1, 2, 3, 4 and 5), we placed the esophageal stent first, followed by the tracheobronchial stent, because the symptoms of esophageal stenosis or fistula were more conspicuous than those of airway stenosis. In the remaining 3 patients (patients 6, 7 and 8), who had both esophageal and tracheobronchial stenoses before the stent therapy, we placed the tracheobronchial stents first, followed by the esophageal stents, because, if the esophageal stents had been placed first, we would have risked deterioration of the tracheobronchial stenoses from compression by the esophageal stent. All patients except one (patient 7) underwent two-staged stent placements. Patient 7 underwent the placement of an airway stent followed by an esophageal stent on the same day. The mean periods between the diagnosis of the disease and stent placements were 9 months (range, from 2–20 months) for the airway stents and 6 months (range, from 0 –18 months) for the esophageal stents. To assess the results of stent therapy, we classified tracheobronchial stenosis into 4 grades as follows: grade 0, few complaints; grade 1, medium-grade stenosis; grade 2, high-grade stenosis without cyanosis; and grade 3, subtotal stenosis with cyanosis. Esophageal stenoses were classified into 4 grades as follows: grade 0, a patient can consume any kind of food; grade 1, a patient can consume soft foods without the nutritional support; grade 2, a patient can consume liquid food without the nutritional support; and grade 3, a patient needs some nutritional supports by either tube feeding or intravenous hyperalimentation because of an insufficient oral intake.
M ⫽ male;
MS ⫽ metallic stent.
Results Immediate improvement of grades of both respiratory symptoms and dysphagia was achieved in all patients after stent placements (Table 2). The symptoms of tracheobronchial stenoses did not recur in any of our cases until the patients’ death. The symptoms of esophagotracheobronchial fistulae also improved after the placement of esophageal covered metallic stent. The mean period of taking meals without the nutritional support after esophageal stent placements was 5 months (range, from 2–16 months). The major complications were: (1) respiratory distress in 2 patients because of severe tracheal stenosis from compression by a single esophageal stent; (2) occurrence of a fistulae after double-stent placement in 2 patients; (3) occurrence of a fistula at the time of placement of a Dumon stent in 1 patient; and (4) recurrence of fistula symptoms due to growth of fistula after doublestent placement in 2 patients. In the next, we present these complications in detail. Patients 1 and 2 (see Table 1) had both esophageal and tracheal stenoses before the stent therapy, but no respiratory symptoms (Fig 1). They complained of severe dyspnea immediately after the placement of their esophageal stents: these symptoms were caused by deterioration of the tracheal stenosis related to compression by the esophageal stent. We therefore removed their esophageal stents promptly, resulting in immediate relief of symptoms in both patients. We then placed tracheal stents followed by another esophageal stents. Patients 1 and 2 suffered new esophagotracheal fistulae at the collateral to the stent 5 and 4 weeks, respectively, after placement of the esophageal and tracheal stents (Figs 2A and B). Because a noncovered metallic stent or silicone stents were used in the esophagus in these patients, they complained of fistula symptoms. Because covered metallic stents were not commercially available at the time of treatment, we managed these patients by stopping oral feeding and switching to tube feeding or intravenous hyperalimentation. Patient 4 suffered an esophagotracheal fistula at the
Ann Thorac Surg 2000;70:1803–7
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
1805
Table 2. Results of Stenting After Esophageal and Tracheobronchial Stenoses
Patient No.
Grade of Airway Stenosis
Grade of Es. Stenosis
Fistula
Before
After
Before
After
Before
After
1
3
0
3
1
⫺
⫹
2
3
0
3
1
⫺
⫹
3 4
1 3
0 0
3 3
1 1
⫺ ⫺
⫹ ⫹
5
3
1
2
1
⫹
⫹⫹
6 7
2 2
0 0
3 2
1 1
⫹ ⫹
⫹ ⫹⫹
8
3
0
3
1
⫹
⫹
Es ⫽ esophageal;
IVH ⫽ intravenous hyperalimentation;
Complications of Stenting Asphyxia, new fistula Asphyxia, new fistula ... New fistula without symptoms Reopened fistula ... Reopened fistula ...
Treatment for Complications Remove MS, Tube feeding Remove MS, IVH ... ...
Additional covered MS ... Additional covered MS ...
Period of Taking a Meal After Es. Stent (Months)
Prognosis After Double Stenting
2
96 days Dead
16
74 days Dead
5 4
38 days Dead 64 days Dead
4
73 days Dead
2 2
76 days Dead 82 days Dead
7
212 days Alive
MS ⫽ metallic stent.
time of Dumon-stent placement because of its procedure. He had received a covered metallic stent for esophageal cancer 2 months previously. The patient, however, did not complain of fistula symptoms, as the fistula was blocked by the covered metallic stent that had been previously placed in the esophagus (Fig 3). Patients 5 and 7 had recurring symptoms of fistulae 1 and 10 weeks, respectively, after the placement of stents in both the esophagus and tracheobronchus. These recurring symptoms were caused by growth of the preexisting fistulae: pressure from both the stents caused extensive necrosis of the esophageal and tracheobronchial walls. Placement of additional esophageal covered metallic stents improved the fistula symptoms in both patients, who became able to take a meal without nutritional support. Seven patients died of their tumors from 38 to 96 days (mean: 72 days) after placement of the stents in both the esophagus and tracheobronchus. Patient 6 died of massive bleeding from the tumor 76 days after the double stenting. The remaining patient is now still alive 211 days after the double stenting without dysphagia or dyspnea.
Fig 1. CT showing stenoses of both the esophagus and trachea.
Comment Although our study showed that double stenting for esophageal and tracheobronchial stenoses effectively relieved symptoms, we discovered 2 main complications: (1) in patients with both esophageal and tracheal stenoses, the tracheal stenoses can deteriorate if they are compressed by the esophageal stents; and (2) with double stents, there is a high risk of fistula occurring or growing because of necrosis of both the esophageal and tracheobronchial walls from the pressure of the stents. Two of our patients suffered the first complication, that is, deterioration of tracheal stenosis because of compression by the esophageal stent. They had both esophageal and tracheal stenoses, but did not complain of respiratory symptoms before placement of the esophageal stent. Placement of the esophageal stents compressed the trachea, resulting in severe tracheal stenosis and necessitating prompt removal of the esophageal stents. There have been no reports regarding this complication, but we recommend that: (1) tracheal stenoses should be examined by CT and bronchoscopy before placement of esophageal stents; and (2) for patients with both the esophageal and tracheal stenoses, the tracheal stent should be placed before the esophageal stent even if the patients does not have respiratory symptoms. Five of our patients suffered the second complication listed above, that is, development or growth of esophagotracheobronchial fistulae after placement of both the esophageal and tracheobronchial stents. Of these, 4 patients underwent esophageal stent placement first and the remaining 1 underwent the tracheobronchial stent placement first. None of the patients suffered development or growth of fistulae during single-stent placement in the esophagus or tracheobronchus: these problems occurred after double stenting. In all the patients except 1, we used self-expandable metallic stents for esophageal
1806
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
Ann Thorac Surg 2000;70:1803–7
into the esophagus at first to block the fistula, whereas additional tracheobronchial stents should not be used to prevent the growth of the fistula as long as the tracheobronchial stenosis does not become more severe. Maier and associates reported that 4 of 11 patients (36.3%) suffered esophageal rupture during single-stent placement for esophagus, and their patients had undergone excessive endoluminal pretreatment, such as photodynamic therapy followed by endoluminal high-dose brachyradiotherapy, before stent therapy, making the esophageal wall prone to be ruptured [8]. Our patients underwent chemoradiotherapy before stent therapy, but they did not receive locally excessive endoluminal treatment such as the Maier’s method, which could be a cause of no problems for fistula occurring or growing during single-stent placement in the present study. We recommend the use of a covered metallic stent in esophageal cancer invading the tracheobronchus, even for patients without esophagotracheobronchial fistulae, because there is a high risk of fistula occurring in the future, especially if double stents need to be placed. Three of our patients (patients 4, 6 and 8) did not complain of fistula symptoms despite having fistulae, because they had been pretreated with covered metallic stents. Because a covered metallic stent can prevent or
Fig 2. (A) CT showing an esophagotracheal fistula beside the stents (arrow). (B) Chest X-ray showing a silicone stent in the esophagus (arrows) and a dynamic stent in the trachea.
stenoses, which gradually expand to its final outer diameter of 22 mm after placement, without extensive compression to the esophageal wall [6, 7]. Although the Dumon stent and dynamic stent are more rigid than the self-expandable metallic stent, their diameters are selected to match the lumen of tracheobronchus, avoiding excessive compression of the tracheobronchial wall. However, the pressure from both the stents placed at the esophagus and tracheobronchus can cause extensive necrosis of the walls, leading to the development or growth of fistulae. For esophageal cancer with fistula but with little tracheobronchial stenosis, we therefore recommend that only a covered metallic stent should be placed
Fig 3. Dumon stent and covered metallic stent placed in the trachea and esophagus, respectively.
Ann Thorac Surg 2000;70:1803–7
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
minimize fistula symptoms [9, 10], it is the treatment of choice for esophageal cancer invading the tracheobronchus, even for patients without fistulae. In conclusion, although the placement of stents in both the esophagus and tracheobronchus can effectively relieve symptoms of dysphagia and airway stenosis, it should be kept in mind that: (1) for patients with both the esophageal and tracheobronchial stenoses, the tracheobronchial stent should be placed first to prevent worsening of the airway stenosis from compression by the esophageal stent; and (2) a covered metallic stent should be used for esophageal tumors involving tracheobronchus even for patients without the fistulae, because the double stents increases the risk of development or growth of esophagotracheobronchial fistulae.
References 1. Wallace MJ, Charnsangavej D, Ogawa K, Carrasco CH, Wright KC, McKenna R, McMurtrey M, Gianturco C. Tracheobronchial tree: expandable metallic stents used in experimental and clinical applications. Work in progress. Radiology 1986;158:309 –12. 2. Nomori H, Kobayashi R, Kodera K, Morinaga S, Ogawa K. Indications for an expandable metallic stent for tracheobronchial stenosis. Ann Thorac Surg 1993;56:1324 – 8.
1807
3. Dumon JF. A dedicated tracheobronchial stent. Chest 1990; 97:328 –32. 4. Nomori H, Horio H, Suemasu K. Dumon stent placement via endotracheal tube. Chest 1999;115:582–3. 5. Freitag L, Eicker R, Linz B, Greschuchna D. Theoretical and experimental basis for the development of a dynamic airway stent. Eur Respir J 1994;7:2038 – 45. 6. Ell C, Hochberger J, May A, Fleig WE, Hahn EG. Coated and uncoated self-expanding metal stents for malignant stenosis in the upper GI tract: preliminary clinical experiences with wall stents. Am J Gastroenterol 1994;89:1496 – 500. 7. Knyrim K, Wanger HJ, Bethge N, Keymling M, Vakil N. A controlled trial of an expansible metal stent for palliation of esophageal obstruction due to inoperable cancer. N Engl J Med 1993;329:1302–7. 8. Maier A, Pinter H, Friehs GB, Renner H, Smolle-Ju¨ttner FM. Self-expandable coated stent after intraluminal treatment of esophageal cancer: a risky procedure? Ann Thorac Surg 1999;67:781– 4. 9. Wong K, Goldstraw P. Role of covered esophageal stents in malignant esophagorespiratory fistula. Ann Thorac Surg 1995;60:199 –200. 10. Wu WC, Katon RM, Saxon RR, Barton RE, Uchida BT, Keller FS, Rosch J. Silicone-covered self expanding metallic stents for the palliation of malignant esophageal obstruction and esophagorespiratory fistulas: experience in 32 patients and a review of the literature. Gastrointest Endosc 1994;40: 22–33.
INVITED COMMENTARY The introduction of self-expanding wire mesh stents has greatly simplified management of uncomplicated malignant esophageal and airway stenoses. However, when malignant obstructions become extensive enough to involve both the tracheobronchial tree and the esophagus, or when a malignant esophageal-airway fistula develops, palliation of stridor, dysphagia, and aspiration is challenging. With their multiple case reports, Nomori and coauthors illustrated the types of complications and potential benefits that result from combined airway and esophageal stenting. Most patients experienced improvement of both airway and esophageal symptoms after double stenting, albeit for a relatively short time, given that the average life expectancy after the second stent placement was just over 2 months. Of concern is the high likelihood that sandwiching the membranous trachea and the esophagus between two stents will result in tissue necrosis and the development or worsening of esophageal-airway fistulas, which occurred in five of the eight patients in this series. In less skilled hands, these complications might have been catastrophic. Although it appears to have been of benefit in most patients treated by Nomori and colleagues, the suggestion that placement of additional stents is appropriate for managing iatrogenic fistulas associated with double stenting is counterintuitive. Those involved in the man-
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
agement of these challenging patients may wish to consider other alternatives. Initial placement of stents that have smaller diameters may provide sufficient palliation of symptoms without causing excess pressure on surrounding tissues. When tumors are extremely bulky and combined airway and esophageal compromise is likely to develop in the future, a longer length of esophageal stent than would normally be chosen may prevent any fistula that may develop from becoming symptomatic. Use of the more dynamic self-expanding metal stents rather than silicone or plastic stents in the airway may reduce the propensity for fistula formation. Avoiding doublestent placement except as a last resort, particularly in people with preexisting airway stents, will also limit the likelihood of fistula formation. Ultimately, however, the development of new devices will be necessary to palliate these difficult and rapidly fatal problems. Mark K. Ferguson, MD Department of Surgery The University of Chicago Medical Center 5841 South Maryland Ave MC 5035 Chicago, IL 60637 e-mail: [email protected].
Ann Thorac Surg 2000;70:1807
• 0003-4975/00/$20.00 PII S0003-4975(00)02041-5
Background. We examined the complications and outcomes of placing stents for both esophageal and tracheobronchial stenoses. Methods. We placed stents for both esophageal and tracheobronchial stenoses in 8 patients (7 with esophageal cancer and 1 with lung cancer). Covered or noncovered metallic stents were used for the esophageal stenoses, except in 1 patient treated with a silicone stent. Silicone stents were used for the tracheobronchial stenoses. The grades of esophageal and tracheobronchial stenoses were scored. Results. All patients experienced improvement of grades of both dysphagia and respiratory symptoms after stent therapy. The complications were: (1) 2 patients suffered respiratory distress after placement of the esophageal stent because of compression of the trachea by the stent; and (2) 3 patients developed new esophago-
tracheobronchial fistulae, and 2 patients had recurring fistula symptoms because of growth of preexisting fistulae after the stent placement, which were caused by pressure from the 2 stents. Despite the fistulae, the 5 patients treated with covered metallic stents did not complain of fistula symptoms, but 2 patients treated with noncovered metallic or silicone stents did complain. Conclusions. For patients with both esophageal and tracheobronchial stenoses, a stent should be introduced into the tracheobronchus first. Because placement of stents in both the esophagus and tracheobronchus has a high risk of enlargement of the fistula, a covered metallic stent is preferable for esophageal cancer involving the tracheobronchus.
R
esophageal cancer in 7 patients and lung cancer in 1. The mean length of esophageal stenoses in the esophageal cancer patients was 5.7 cm (range: 3–9). The sites of airway stenoses were the trachea in 5 patients, the left main bronchus in 2, and the right main bronchus in 1. All patients underwent chemoradiotherapy before the stent placements. In addition, one patient (patient 1) underwent also radiotherapy after the stent placement. Four patients (patient 5, 6, 7 and 8) already had fistulae before stent therapy. For tracheobronchial stenosis, we used silicone stents but not metallic stents to prevent a recurrence of stenosis due to tumor growth between the wires of the metallic stents [2]. Six Dumon stents (Novatech Co, Aubagne, France) were placed in 6 patients and 2 dynamic stents (Willy Ru¨sch AG, Kernen, Germany) in 2. Dumon stents were placed via an endotracheal tube under general anesthesia as described previously [4]. The dynamic stents were placed using specialized forceps under general anesthesia. The esophageal metallic stents were placed using the regular method under local anesthesia [5–7]. For esophageal stenosis, we generally used metallic stents but not silicone stents, because silicone stents tend to have a higher risk of esophageal injury and are also easily packed with food. Nine covered selfexpandable metallic stents (Covered Ultraflex Esophageal Stent System, Microinvasive, Boston Scientific Co, Boston, MA) were placed in 6 patients, 2 noncovered Ultraflex metallic stents in 2 patients, and 1 silicone stent
ecent advances in stent therapy have increased the quality of life of patients with inoperable cancer. Tracheobronchial stenoses have been treated by expandable metallic stent [1, 2] and silicone stent such as Dumon or dynamic stents [3–5]. Esophageal stenoses have classically been treated with plastic or silicone stent, but self-expandable metallic stent have recently been introduced for this purpose [5, 6]. Patients with advanced esophageal cancer often have respiratory distress as a result of tracheobronchial invasion, and this necessitates stent therapy not only for esophageal stenosis but also for tracheobronchial stenosis. In this report we examine the outcomes of stent therapy for both esophageal and tracheobronchial stenoses in 8 patients and recommend some management techniques that can be used to prevent complications.
Material and Methods From 1991 to 1999, we performed stent therapy for 28 patients with esophageal stenoses and 45 patients with tracheobronchial stenoses. Of these, 8 patients received stent therapy for both esophageal and tracheobronchial stenoses (Table 1). The causes of the stenoses were Accepted for publication May 14, 2000. Address reprint requests to Dr Nomori, Department of Surgery, Saiseikai Central Hospital, 1-4-17 Mita, Minato-ku, Tokyo 108-0073, Japan; e-mail: [email protected].
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2000;70:1803–7) © 2000 by The Society of Thoracic Surgeons
0003-4975/00/$20.00 PII S0003-4975(00)02042-7
1804
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
Ann Thorac Surg 2000;70:1803–7
Table 1. Patients Characteristics and Treatments for Esophageal and Airway Stenosis Airway Stent Patient No.
Age/ Gender
Cause of Stenoses
Site of Airway Stenoses
Fistula Before Stenting
Type
Time (months)a
Esophageal Stent Type (Numbers)
1 2
56/F 67/M
Es. Ca Es. Ca
Trachea Trachea
⫺ ⫺
Dynamic Dynamic
3 14
3 4 5
65/M 65/M 45/M
Es. Ca Es. Ca Es. Ca
Main Br. Trachea Trachea
⫺ ⫺ ⫹
Dumon Dumon Dumon
5 20 12
6 7
42/F 65/M
Es. Ca Es. Ca
Main Br. Main Br.
⫹ ⫹
Dumon Dumon
2 4
Covered MS (1) Covered MS (2)
8
76/M
Lung Ca
Trachea
⫹
Dumon
12
Covered MS (1)
a
Noncovered MS (1) Silicone stent (1) Noncovered MS (1) Covered MS (1) Covered MS (1) Covered MS (2)
Time (months)a 3 0 5 0 18 9 12 2 4 6 12
Time means a period between the diagnosis of the diseases and the stent placement.
Br ⫽ bronchus;
Ca ⫽ cancer;
Es ⫽ esophageal;
F ⫽ female;
(Flexible Artificial Esophagus, Sumitomo Bakelite Co, Ltd, Tokyo, Japan) in 1. The main purposes for using these esophageal stents were to dilate the stenoses in 6 patients (patients 1, 2, 3, 4, 6 and 8) and to block the esophagotracheobronchial fistula in the 2 (patients 5 and 7). A noncovered metallic stent or a silicone stent had to be used in 2 patients (patients 1 and 2) because a covered metallic stent was not commercially available in Japan at the time of treatment. In 5 patients (patients 1, 2, 3, 4 and 5), we placed the esophageal stent first, followed by the tracheobronchial stent, because the symptoms of esophageal stenosis or fistula were more conspicuous than those of airway stenosis. In the remaining 3 patients (patients 6, 7 and 8), who had both esophageal and tracheobronchial stenoses before the stent therapy, we placed the tracheobronchial stents first, followed by the esophageal stents, because, if the esophageal stents had been placed first, we would have risked deterioration of the tracheobronchial stenoses from compression by the esophageal stent. All patients except one (patient 7) underwent two-staged stent placements. Patient 7 underwent the placement of an airway stent followed by an esophageal stent on the same day. The mean periods between the diagnosis of the disease and stent placements were 9 months (range, from 2–20 months) for the airway stents and 6 months (range, from 0 –18 months) for the esophageal stents. To assess the results of stent therapy, we classified tracheobronchial stenosis into 4 grades as follows: grade 0, few complaints; grade 1, medium-grade stenosis; grade 2, high-grade stenosis without cyanosis; and grade 3, subtotal stenosis with cyanosis. Esophageal stenoses were classified into 4 grades as follows: grade 0, a patient can consume any kind of food; grade 1, a patient can consume soft foods without the nutritional support; grade 2, a patient can consume liquid food without the nutritional support; and grade 3, a patient needs some nutritional supports by either tube feeding or intravenous hyperalimentation because of an insufficient oral intake.
M ⫽ male;
MS ⫽ metallic stent.
Results Immediate improvement of grades of both respiratory symptoms and dysphagia was achieved in all patients after stent placements (Table 2). The symptoms of tracheobronchial stenoses did not recur in any of our cases until the patients’ death. The symptoms of esophagotracheobronchial fistulae also improved after the placement of esophageal covered metallic stent. The mean period of taking meals without the nutritional support after esophageal stent placements was 5 months (range, from 2–16 months). The major complications were: (1) respiratory distress in 2 patients because of severe tracheal stenosis from compression by a single esophageal stent; (2) occurrence of a fistulae after double-stent placement in 2 patients; (3) occurrence of a fistula at the time of placement of a Dumon stent in 1 patient; and (4) recurrence of fistula symptoms due to growth of fistula after doublestent placement in 2 patients. In the next, we present these complications in detail. Patients 1 and 2 (see Table 1) had both esophageal and tracheal stenoses before the stent therapy, but no respiratory symptoms (Fig 1). They complained of severe dyspnea immediately after the placement of their esophageal stents: these symptoms were caused by deterioration of the tracheal stenosis related to compression by the esophageal stent. We therefore removed their esophageal stents promptly, resulting in immediate relief of symptoms in both patients. We then placed tracheal stents followed by another esophageal stents. Patients 1 and 2 suffered new esophagotracheal fistulae at the collateral to the stent 5 and 4 weeks, respectively, after placement of the esophageal and tracheal stents (Figs 2A and B). Because a noncovered metallic stent or silicone stents were used in the esophagus in these patients, they complained of fistula symptoms. Because covered metallic stents were not commercially available at the time of treatment, we managed these patients by stopping oral feeding and switching to tube feeding or intravenous hyperalimentation. Patient 4 suffered an esophagotracheal fistula at the
Ann Thorac Surg 2000;70:1803–7
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
1805
Table 2. Results of Stenting After Esophageal and Tracheobronchial Stenoses
Patient No.
Grade of Airway Stenosis
Grade of Es. Stenosis
Fistula
Before
After
Before
After
Before
After
1
3
0
3
1
⫺
⫹
2
3
0
3
1
⫺
⫹
3 4
1 3
0 0
3 3
1 1
⫺ ⫺
⫹ ⫹
5
3
1
2
1
⫹
⫹⫹
6 7
2 2
0 0
3 2
1 1
⫹ ⫹
⫹ ⫹⫹
8
3
0
3
1
⫹
⫹
Es ⫽ esophageal;
IVH ⫽ intravenous hyperalimentation;
Complications of Stenting Asphyxia, new fistula Asphyxia, new fistula ... New fistula without symptoms Reopened fistula ... Reopened fistula ...
Treatment for Complications Remove MS, Tube feeding Remove MS, IVH ... ...
Additional covered MS ... Additional covered MS ...
Period of Taking a Meal After Es. Stent (Months)
Prognosis After Double Stenting
2
96 days Dead
16
74 days Dead
5 4
38 days Dead 64 days Dead
4
73 days Dead
2 2
76 days Dead 82 days Dead
7
212 days Alive
MS ⫽ metallic stent.
time of Dumon-stent placement because of its procedure. He had received a covered metallic stent for esophageal cancer 2 months previously. The patient, however, did not complain of fistula symptoms, as the fistula was blocked by the covered metallic stent that had been previously placed in the esophagus (Fig 3). Patients 5 and 7 had recurring symptoms of fistulae 1 and 10 weeks, respectively, after the placement of stents in both the esophagus and tracheobronchus. These recurring symptoms were caused by growth of the preexisting fistulae: pressure from both the stents caused extensive necrosis of the esophageal and tracheobronchial walls. Placement of additional esophageal covered metallic stents improved the fistula symptoms in both patients, who became able to take a meal without nutritional support. Seven patients died of their tumors from 38 to 96 days (mean: 72 days) after placement of the stents in both the esophagus and tracheobronchus. Patient 6 died of massive bleeding from the tumor 76 days after the double stenting. The remaining patient is now still alive 211 days after the double stenting without dysphagia or dyspnea.
Fig 1. CT showing stenoses of both the esophagus and trachea.
Comment Although our study showed that double stenting for esophageal and tracheobronchial stenoses effectively relieved symptoms, we discovered 2 main complications: (1) in patients with both esophageal and tracheal stenoses, the tracheal stenoses can deteriorate if they are compressed by the esophageal stents; and (2) with double stents, there is a high risk of fistula occurring or growing because of necrosis of both the esophageal and tracheobronchial walls from the pressure of the stents. Two of our patients suffered the first complication, that is, deterioration of tracheal stenosis because of compression by the esophageal stent. They had both esophageal and tracheal stenoses, but did not complain of respiratory symptoms before placement of the esophageal stent. Placement of the esophageal stents compressed the trachea, resulting in severe tracheal stenosis and necessitating prompt removal of the esophageal stents. There have been no reports regarding this complication, but we recommend that: (1) tracheal stenoses should be examined by CT and bronchoscopy before placement of esophageal stents; and (2) for patients with both the esophageal and tracheal stenoses, the tracheal stent should be placed before the esophageal stent even if the patients does not have respiratory symptoms. Five of our patients suffered the second complication listed above, that is, development or growth of esophagotracheobronchial fistulae after placement of both the esophageal and tracheobronchial stents. Of these, 4 patients underwent esophageal stent placement first and the remaining 1 underwent the tracheobronchial stent placement first. None of the patients suffered development or growth of fistulae during single-stent placement in the esophagus or tracheobronchus: these problems occurred after double stenting. In all the patients except 1, we used self-expandable metallic stents for esophageal
1806
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
Ann Thorac Surg 2000;70:1803–7
into the esophagus at first to block the fistula, whereas additional tracheobronchial stents should not be used to prevent the growth of the fistula as long as the tracheobronchial stenosis does not become more severe. Maier and associates reported that 4 of 11 patients (36.3%) suffered esophageal rupture during single-stent placement for esophagus, and their patients had undergone excessive endoluminal pretreatment, such as photodynamic therapy followed by endoluminal high-dose brachyradiotherapy, before stent therapy, making the esophageal wall prone to be ruptured [8]. Our patients underwent chemoradiotherapy before stent therapy, but they did not receive locally excessive endoluminal treatment such as the Maier’s method, which could be a cause of no problems for fistula occurring or growing during single-stent placement in the present study. We recommend the use of a covered metallic stent in esophageal cancer invading the tracheobronchus, even for patients without esophagotracheobronchial fistulae, because there is a high risk of fistula occurring in the future, especially if double stents need to be placed. Three of our patients (patients 4, 6 and 8) did not complain of fistula symptoms despite having fistulae, because they had been pretreated with covered metallic stents. Because a covered metallic stent can prevent or
Fig 2. (A) CT showing an esophagotracheal fistula beside the stents (arrow). (B) Chest X-ray showing a silicone stent in the esophagus (arrows) and a dynamic stent in the trachea.
stenoses, which gradually expand to its final outer diameter of 22 mm after placement, without extensive compression to the esophageal wall [6, 7]. Although the Dumon stent and dynamic stent are more rigid than the self-expandable metallic stent, their diameters are selected to match the lumen of tracheobronchus, avoiding excessive compression of the tracheobronchial wall. However, the pressure from both the stents placed at the esophagus and tracheobronchus can cause extensive necrosis of the walls, leading to the development or growth of fistulae. For esophageal cancer with fistula but with little tracheobronchial stenosis, we therefore recommend that only a covered metallic stent should be placed
Fig 3. Dumon stent and covered metallic stent placed in the trachea and esophagus, respectively.
Ann Thorac Surg 2000;70:1803–7
NOMORI ET AL DOUBLE STENTING FOR ESOPHAGEAL AND TRACHEOBRONCHIAL STENOSES
minimize fistula symptoms [9, 10], it is the treatment of choice for esophageal cancer invading the tracheobronchus, even for patients without fistulae. In conclusion, although the placement of stents in both the esophagus and tracheobronchus can effectively relieve symptoms of dysphagia and airway stenosis, it should be kept in mind that: (1) for patients with both the esophageal and tracheobronchial stenoses, the tracheobronchial stent should be placed first to prevent worsening of the airway stenosis from compression by the esophageal stent; and (2) a covered metallic stent should be used for esophageal tumors involving tracheobronchus even for patients without the fistulae, because the double stents increases the risk of development or growth of esophagotracheobronchial fistulae.
References 1. Wallace MJ, Charnsangavej D, Ogawa K, Carrasco CH, Wright KC, McKenna R, McMurtrey M, Gianturco C. Tracheobronchial tree: expandable metallic stents used in experimental and clinical applications. Work in progress. Radiology 1986;158:309 –12. 2. Nomori H, Kobayashi R, Kodera K, Morinaga S, Ogawa K. Indications for an expandable metallic stent for tracheobronchial stenosis. Ann Thorac Surg 1993;56:1324 – 8.
1807
3. Dumon JF. A dedicated tracheobronchial stent. Chest 1990; 97:328 –32. 4. Nomori H, Horio H, Suemasu K. Dumon stent placement via endotracheal tube. Chest 1999;115:582–3. 5. Freitag L, Eicker R, Linz B, Greschuchna D. Theoretical and experimental basis for the development of a dynamic airway stent. Eur Respir J 1994;7:2038 – 45. 6. Ell C, Hochberger J, May A, Fleig WE, Hahn EG. Coated and uncoated self-expanding metal stents for malignant stenosis in the upper GI tract: preliminary clinical experiences with wall stents. Am J Gastroenterol 1994;89:1496 – 500. 7. Knyrim K, Wanger HJ, Bethge N, Keymling M, Vakil N. A controlled trial of an expansible metal stent for palliation of esophageal obstruction due to inoperable cancer. N Engl J Med 1993;329:1302–7. 8. Maier A, Pinter H, Friehs GB, Renner H, Smolle-Ju¨ttner FM. Self-expandable coated stent after intraluminal treatment of esophageal cancer: a risky procedure? Ann Thorac Surg 1999;67:781– 4. 9. Wong K, Goldstraw P. Role of covered esophageal stents in malignant esophagorespiratory fistula. Ann Thorac Surg 1995;60:199 –200. 10. Wu WC, Katon RM, Saxon RR, Barton RE, Uchida BT, Keller FS, Rosch J. Silicone-covered self expanding metallic stents for the palliation of malignant esophageal obstruction and esophagorespiratory fistulas: experience in 32 patients and a review of the literature. Gastrointest Endosc 1994;40: 22–33.
INVITED COMMENTARY The introduction of self-expanding wire mesh stents has greatly simplified management of uncomplicated malignant esophageal and airway stenoses. However, when malignant obstructions become extensive enough to involve both the tracheobronchial tree and the esophagus, or when a malignant esophageal-airway fistula develops, palliation of stridor, dysphagia, and aspiration is challenging. With their multiple case reports, Nomori and coauthors illustrated the types of complications and potential benefits that result from combined airway and esophageal stenting. Most patients experienced improvement of both airway and esophageal symptoms after double stenting, albeit for a relatively short time, given that the average life expectancy after the second stent placement was just over 2 months. Of concern is the high likelihood that sandwiching the membranous trachea and the esophagus between two stents will result in tissue necrosis and the development or worsening of esophageal-airway fistulas, which occurred in five of the eight patients in this series. In less skilled hands, these complications might have been catastrophic. Although it appears to have been of benefit in most patients treated by Nomori and colleagues, the suggestion that placement of additional stents is appropriate for managing iatrogenic fistulas associated with double stenting is counterintuitive. Those involved in the man-
© 2000 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
agement of these challenging patients may wish to consider other alternatives. Initial placement of stents that have smaller diameters may provide sufficient palliation of symptoms without causing excess pressure on surrounding tissues. When tumors are extremely bulky and combined airway and esophageal compromise is likely to develop in the future, a longer length of esophageal stent than would normally be chosen may prevent any fistula that may develop from becoming symptomatic. Use of the more dynamic self-expanding metal stents rather than silicone or plastic stents in the airway may reduce the propensity for fistula formation. Avoiding doublestent placement except as a last resort, particularly in people with preexisting airway stents, will also limit the likelihood of fistula formation. Ultimately, however, the development of new devices will be necessary to palliate these difficult and rapidly fatal problems. Mark K. Ferguson, MD Department of Surgery The University of Chicago Medical Center 5841 South Maryland Ave MC 5035 Chicago, IL 60637 e-mail: [email protected].
Ann Thorac Surg 2000;70:1807
• 0003-4975/00/$20.00 PII S0003-4975(00)02041-5