| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Asan Medical Center, 388-1, Poongnap-dong, Songpa-gu, Seoul 138-736, South Korea.
Received February 9, 2004;
accepted after revision March 18, 2004.
Address correspondence to J. H. Shin
(jhshin{at}amc.seoul.kr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Under fluoroscopic guidance, the removal of 119 esophageal, six gastroduodenal, and five rectal retrievable stents was attempted in 113 patients using a retrieval hook. Indications for stent removal included migration (n = 35), severe pain (n = 23), formation of a new stricture (n = 13), incomplete stent expansion (n = 7), airway compression (n = 2), esophagorespiratory fistula (n = 2), malpositioned stent (n = 1), and hematemesis (n = 1). The remaining 46 stents were electively removed.
RESULTS. Of the 130 stents, 127 (97.7%) were successfully removed despite the following difficulties: untied drawstrings (n = 4), separation of the stent (n = 3), and fracture (n = 2) or disconnection (n = 2) of a retrieval hook. The removal procedure failed in three cases (2.3%). The causes of failure were the inability to place the hook into the migrated stent (n = 2) and a tight stricture above the migrated stent (n = 1). The procedure-related complications included minor (n = 4) and major (n = 1) bleeding and intramural rupture (n = 3). One patient died of major bleeding after removal of an esophageal stent.
CONCLUSION. The stent retrieval hook is useful for removing retrievable esophageal and gastrointestinal stents.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this article, we describe a hook technique to remove specially designed retrievable esophageal and gastrointestinal stents. This technique is useful only for removing the retrievable stents, not for the other commercially available stents. The purpose of our study was to evaluate the safety and efficacy of removing retrievable esophageal and gastrointestinal stents with a retrieval hook.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Esophageal Stent
Using a retrieval hook, we attempted to remove 119 Song stents in 104
patients—72 male and 32 female patients (age range, 1–82 years;
mean age, 54.9 years). These patients had undergone stent placement for
esophageal strictures (n = 100) or for esophagorespiratory fistulas
(n = 4). The causes of esophageal stricture or esophagorespiratory
fistula were malignant in 63 patients and benign in 41 patients. The malignant
causes were inoperable esophageal cancer (n = 50), recurrent
malignant tumor after surgery (gastrectomy [n = 5] or laryngectomy
[n = 3]), inoperable gastric cancer involving the esophagogastric
junction (n = 1), metastasis from lung (n = 2) or rectal
(n = 1) cancer, and unknown (n = 1). The benign causes were
ingestion of corrosive agents (n = 34), anastomotic stricture after
surgery for esophageal atresia (n = 3), gastric cancer (n =
1), small-bowel adenoma (n = 1), tuberculous mediastinitis
(n = 1), and reflux esophagitis (n = 1).
The construction of the Song stent was described in detail by Song et al. [9–11]. Briefly, the Song stent is woven from a single thread of 0.2-mm nitinol wire in a tubular configuration and is covered in polyurethane solution (Biospan, Polymer Technology Group) by a dipping method. Both ends of the stent are 4–6 mm wider in diameter than the body of the stent to prevent migration. Two drawstrings are attached to the upper inner margin of the stent to facilitate removal (Fig. 1). The esophageal stents were 8–18 mm (mean, 16.0 mm) in diameter and 3–16 cm (mean, 9.7 cm) in length.
|
In our patients, the stents were placed in the cervical (n = 8), upper thoracic (n = 18), middle thoracic (n = 51), lower thoracic (n = 31), and entire thoracic (n = 5) esophagus and at the site of an esophagojejunostomy (n = 6). Of the 119 stents, 75 were removed because of complications such as stent migration (total, n = 27; upward, n = 15; and downward, n = 12), severe pain requiring narcotic analgesics (n = 23), formation of a new stricture above or below the stent (n = 13), incomplete expansion of the stent (n = 7), airway compression by the stent (n = 2), development of an esophagorespiratory fistula at the stented segment (n = 2), and hematemesis (n = 1). Of the 44 electively removed stents, 18 stents placed for corrosive esophageal stricture were removed 4–8 weeks after placement. The remaining 26 stents placed for malignant stricture (esophageal cancer [n = 24], recurrent gastric cancer [n = 1], or metastasis [n = 1]) were removed after 3–6 weeks of radiation therapy to prevent delayed complications. The times of elective stent removal were determined by referring to previous reports [7, 10, 11]. The mean interval between stent placement and removal was 37.8 days (range, 0–214 days).
Gastroduodenal Stent
Removals of six Song gastroduodenal stents were attempted in six patients
(five men and one woman; mean age, 49.3 years) who had undergone stent
placement for a malignant (n = 4) or a benign (n = 2)
stricture. The causes of the malignant strictures were recurrent gastric
cancer after surgery (n = 3) and metastasis from biliary cancer
(n = 1). The causes of the benign strictures were anastomotic
strictures that developed after surgery for superior mesenteric artery
syndrome (n = 1) and gastric cancer (n = 1). The stents were
16 mm in diameter and 8–16 cm (mean, 10 cm) in length. The locations of
the stents were at gastrojejunostomy (n = 4) or duodenojejunostomy
(n = 1) sites and at the antrum of the stomach (n = 1).
The stents were removed using a retrieval hook, either electively (n = 1) or because of complications (n = 5) that included stent migration (n = 4) and malpositioning (n = 1). The electively removed stent placed for a benign stricture at the duodenojejunostomy site was removed 8 weeks after placement. The mean interval between stent placement and removal was 28.2 days (range, 0–86 days).
Rectal Stent
Five Song rectal stents were removed from three patients (all men; ages,
59, 59, and 64 years) who had undergone stent placement for a malignant
(n = 2) or a benign (n = 1) rectal stricture. Malignant
strictures were caused by inoperable (n = 1) or recurrent rectal
cancer after low anterior resection (n = 1). The remaining patient
had a benign anastomotic stricture after low anterior resection for rectal
cancer. The stents were 24–32 mm in diameter and 9–13 cm in
length.
The indications for stent removal were distal stent migration (n = 4) in all malignant strictures. The one stent placed for benign anastomotic stricture was electively removed 8 weeks after placement. The mean interval between stent placement and removal was 23.6 days (range, 2–57 days).
Stent Retrieval Set and Removal Technique
Detailed descriptions of stent retrieval devices have been reported
previously by Song et al.
[9–11].
Briefly, the stent retrieval set consisted of a 13-French sheath, a 10-French
dilator, a hookwire (all, Taewoong) and a 0.035-inch guidewire (Terumo). The
distal end of the hookwire was constructed in a question-mark configuration to
hook the drawstring of the stent. The distal 20-mm portion of the hookwire was
positioned at approximately a 30° angle to the axis
(Fig. 2). The length of the
wire was 70 cm for removal of esophageal and rectal stents and 110 cm for
removal of gastroduodenal stents.
|
Topical anesthesia of the pharynx was achieved with an aerosol spray (atomizer, DeVilbiss). Agents for sedation or general anesthesia were not used. Contrast-enhanced studies were not performed before stent removal. Each stent was removed under fluoroscopic guidance with the patient in the supine position using a retrieval set according to the description given in a previous report [11]. We used three types of techniques for stent removal—the standard, proximal mesh, and eversion techniques. In the standard technique, the hookwire was manipulated to grasp the drawstring. Then the hookwire was withdrawn through the sheath to collapse the proximal stent when it reached the sheath tip. The sheath, hookwire, and the stent were then pulled out of the organs (Fig. 3A, 3B, 3C, 3D, 3E). In cases in which the hook failed to grasp the drawstring but grasped the proximal third of the mesh of the stent, we removed the stent by pulling it out in its expanded state (proximal mesh technique, Fig. 4A, 4B). In cases in which the hook failed to grasp the drawstring or proximal mesh of the stent, we grasped the distal mesh of the stent and pulled it out of the organs. The stent invaginated into itself and was removed in an everted state (eversion technique, Fig. 5A, 5B). To detect the complications of stent removal, we performed a barium study immediately after the stent removal and again at 2- and 4-week follow-up after the procedure.
|
|
|
|
|
|
|
|
|
Analysis
The overall success rate, the causes of technical failure, and the
complications were analyzed. Successful stent removal was defined as the
complete removal of the stent under fluoroscopic guidance using only the stent
retrieval set. Major bleeding was defined as bleeding that required treatment
for cessation, whereas minor bleeding was defined as bleeding that
spontaneously ceased without any treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
During the removal of four stents, the hook was fractured at the bending point while being pulled out of the esophagus. The separated distal parts of the hookwire remained on the drawstring without migrating. The stents and separated distal parts of the hook could be removed together at a subsequent attempted stent removal with the use of another hook.
In another four stents, the drawstrings became untied during the initial attempt of stent removal. These stents were removed using the proximal mesh (n = 3) or eversion (n = 1) techniques.
Three stents split in two at the distal third of the stent, leaving a segment of the stent in the esophagus that was subsequently removed during another attempted stent removal.
During the removal of two completely migrated stents into the stomach, introduction of a guidewire into the stent lumen became difficult because its proximal end abutted the gastric fundus. To change the position of the stents, we inflated the stomach with air while the patient was in a prone position. After the stent moved into body of the stomach, the guidewire could be introduced through the stent lumen, and the stent was successfully removed using the standard technique (Fig. 6A, 6B, 6C).
|
|
|
In two patients in whom the stents were removed because of the formation of a new fistula, the esophageal strictures had improved at the time of stent removal due to radiation therapy, and the new fistula could not be occluded by additional stent placement to the existing 16-mm stents. Therefore, we treated the fistula by replacing the initially placed stents with larger ones (18 mm in diameter).
In one patient in whom the stent was removed because of recurrent and massive hematemesis, the hematemesis did not recur after stent removal.
Stent removal failed in one patient. We attempted removal when the stent migrated into the stomach 25 months after its placement. The stent was successfully pulled out into the mid esophagus using the proximal mesh technique. However, a tight stricture in the upper esophagus prevented further removal of the stent, and it had to be removed surgically.
After removal of four of the stents, minor bleeding developed that spontaneously resolved within 1 hr. Major bleeding occurred in one patient in whom we tried to remove the stent using the standard technique. However, a small polyp that was between the proximal end of the stent and the introducer sheath was removed together with the stent. Active bleeding developed immediately after stent removal. The patient refused further treatment and died of bleeding 2 days later.
Esophagography performed after stent removal revealed intramural ruptures in two patients that were found to have spontaneously resolved on the 2-week follow-up esophagogram.
After stent removal, 55 of 63 patients with malignant strictures underwent another stent placement (n = 43) and gastrostomy (n = 12). The remaining eight patients refused further treatment. Forty-nine patients died of advanced disease 2–35 months (mean, 16.5 months) after stent removal. Thirty of 41 patients with benign strictures underwent balloon dilatation (n = 19), another stent placement (n = 7), or surgery (n = 4) for recurrent strictures. The remaining 11 patients required no further treatment because of improved dysphagia. Eight patients died of unrelated causes 8–67 months (mean, 22.3 months) after stent removal.
Gastroduodenal Stent
Four (66.7%) of the six Song gastroduodenal stents were successfully
removed using the standard (n = 2), proximal mesh (n = 1),
or eversion (n = 1) technique.
In the removal of one stent that had been placed for benign duodenojejunostomy stricture, another stricture in the gastric antrum prevented removal, resulting in the fracture of the hookwire at its bending point. After balloon dilatation of the stricture, the stent and fractured distal hookwire were successfully removed using the standard technique.
In two patients who had undergone proximal gastrectomy, the migrated stents were vertically located in the remaining gastric antrum. In theses cases, a guidewire could easily be introduced through the stent lumen, and the migrated stents were removed using the proximal mesh technique in one patient and the eversion technique in the other. We easily removed one malpositioned stent using the standard technique, and immediately after the stent removal, a new stent was placed in the proper position.
In two failed removal procedures, the stent had migrated into the gastric fundus. The end of the migrated stent faced in an unfavorable direction in the gastric fundus, and despite air inflation of the stomach and a change in the patient's position, the stent would not move. The guidewires could not be introduced through the stent lumen. Therefore, these stents were removed endoscopically.
Two of four stents removed with a hook were bloodstained. However, active bleeding did not occur during or after the procedure. In both cases, a barium study performed immediately after stent removal showed no complications such as a fistula or leakage of contrast material.
After stent removal, four patients with malignant stricture underwent another stent placement and died of advanced disease 1–11 months after stent removal. Two patients with benign strictures underwent balloon dilatation for recurrent stricture and were alive and without obstructive symptoms 13 and 25 months after stent removal.
Rectal Stent
All five rectal stents were successfully removed using the standard
(n = 4) (Fig. 7A,
7B) or proximal mesh
(n = 1) technique. In one patient, an intramural rupture was seen on
barium study performed immediately after the stent removal. However, the
intramural rupture produced no clinical symptoms and on the follow-up study
performed 2 weeks later was found to have spontaneously resolved.
|
|
After stent removal, two patients with rectal cancer underwent another stent placement and died of advanced disease 6 and 8 months later. One patient with a benign anastomotic stricture underwent balloon dilatation for a recurrent stricture and was alive and without obstructive symptoms 38 months after stent removal.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In previous reports, the indication for removal of esophageal and gastrointestinal stents has been limited to stent migration or malpositioning [8, 14–16]. Recently, Low and Kozarek [15] suggested other indications for stent removal such as overgrowth of granulation tissue, diskitis, meningitis, and epidural abscess. In our study, indications for stent removal included severe pain, formation of a new stricture, incomplete expansion of the stent, airway compression, development of an esophagorespiratory fistula, and hematemesis.
Cwikiel et al. [17] reported a 2% rate for the formation of new strictures in patients who underwent stent placement for malignant strictures. This rate is less than that in patients who undergo stent placement for benign strictures [9–11]. Song et al. [11] reported that formation of a new stricture eventually causes recurrence of dysphagia and necessitates further treatment unless the stent is removed. In addition, some authors have reported that new strictures were found to have improved at follow-up after stent removal [10]. Therefore, we removed the stents when new strictures developed in patients with benign strictures. We also removed the stents when new strictures occurred in patients with malignant strictures in the upper esophagus because additional stent placement to cover the new strictures was difficult.
For benign diseases, an esophageal or gastrointestinal stent is generally placed only temporarily [9, 10, 12, 15]; it would be ideal to electively remove a stent after it is no longer needed. Some authors have reported that removal of esophageal stents 3–10 weeks after placement resulted in complete resolution of an esophagorespiratory fistula [12, 15]. The optimal time to remove a stent placed to treat a benign esophageal stricture has been suggested as 4–8 weeks after placement [11]. We electively removed the stent 6–8 weeks after placement for benign disease. However, Song et al. [10] reported that esophageal strictures recurred in three of seven patients with benign strictures in whom the stents were removed 8 weeks after placement. Therefore, we believe that the optimal time for stent removal is still undetermined.
In our study, the patients who underwent radiation therapy after esophageal stent placement were also candidates for elective stent removal. We based this decision on a previous report that stated that stent removal 4–6 weeks after initiation of radiation therapy in patients with esophagogastric cancers could be an alternative method of palliation not only because the incidence of delayed complications could be reduced but also because abatement of dysphagia was apparent within this period [7].
So far, most reports on the nonsurgical removal of esophageal stents have been descriptions of individual cases [8, 12–14]. Recently, Low and Kozarek [15] reported their experience of endoscopic removal of Ultraflex (Microvasive Endoscopy, Boston Scientific) or Z-esophageal stents (Wilson-Cook Medical) in six patients. The Z-stent was removed by grasping the proximal edge of the stent with a polyp snare. The Ultraflex stent was removed by grasping its distal edge with a rat-toothed forceps and invaginating it into itself. These techniques are similar to the proximal mesh and eversion techniques of our study, respectively. Low and Kozarek reported that all the removal procedures were successful and were without complication. However, we think that these methods have the potential to injure the esophageal wall, and the efficacy and safety of the endoscopic techniques should be confirmed by larger series study.
We removed the stents using a retrieval hook. This device is designed to remove the retrievable Song stent. Our study indicated that removal of a Song stent with a retrieval hook is both feasible and effective. The overall technical success rate was 97.7% (127/130 stents). Technical failure occurred in the removal of two gastroduodenal stents that had migrated upward into the stomach. In these cases, it was difficult to introduce a guidewire through the stent. In five downwardly migrated esophageal stents, a guidewire was easily introduced through the stent with or without inflation of the stomach or a change in the position of the patient. We think that the difference is due to the anatomy of the stomach. The patients with downwardly migrated esophageal stents had normal gastric anatomy. Therefore, by means of simple manipulation, such as air inflation and position change, we could move the stents into the gastric body or antrum in which they lay in a relatively favorable direction for introduction of a guidewire into the lumen. However, because the two patients in whom stent removal failed had undergone subtotal gastrectomy, we could not change the direction of the stent that would abut on the wall of the remnant gastric fundus. Therefore, we could not introduce the guidewire into the stent lumen. We believe that in these cases, endoscopic stent removal should be considered.
The overall complication rate of our study was 6.3% (8/127 removed stents). However, most complications were minor and spontaneously resolved. One patient died of major bleeding after an inadvertent polypectomy during the stent removal. In this patient, the proximal end of the stent incompletely collapsed, suggesting that granulation tissue or a polyp was held within the proximal end of the stent. Although we could not detect the polyp on the esophagogram acquired before the stent removal in this patient, meticulous esophagography and, if available, endoscopic examination would be essential to completely avoid this unexpected complication.
In conclusion, removal of retrievable esophageal and gastrointestinal stents using a retrieval hook is effective and safe. It is a useful method with which to solve various complications caused by these stents.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
