| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
State of the Art |
1 From the Departments of Radiology and Surgery, University of North Carolina School of Medicine, Chapel Hill, NC 27599 (M.A.M.); the Department of Radiology, University of Alabama School of Medicine, Birmingham (R.E.K.); and the Division of Gastroenterology and Hepatology, Mayo School of Medicine, Rochester, Minn (T.H.B.). Received April 1, 1999; revision requested June 14; revision received July 12; accepted July 30. Address correspondence to M.A.M.
Abstract
Metallic stents are currently an established component of the endoluminal treatment of stenoses within the blood vessels, bile ducts, esophagus, trachea, and bronchi. With the development of newer stent designs and delivery systems and the general momentum toward minimally invasive therapies, metallic stent placement has expanded into the nonsurgical therapy for gastroduodenal and colorectal obstructions. The use of metallic stents within the stomach, duodenum, or colon is intended not to be curative but to provide nonsurgical palliation for the symptoms of gastric or colonic obstruction. This palliation may be intended for the life of the patient in the case of unresectable disease or as a temporizing procedure prior to a definitive surgical procedure. In the latter clinical scenario, the benefits of a minimally invasive intestinal decompression procedure include (a) quick and noninvasive relief of the intestinal obstruction in an acutely ill patient that obviates a more extensive procedure; (b) allowance of time to improve a patient's overall medical condition and thus to allow a patient to better tolerate the definitive surgical procedure; and (c) reduction of the complexity of the definitive procedure by eliminating the need for staged procedures and allowing the definitive procedure to be performed at one setting.
Index terms: Gastrointestinal tract, abnormalities, 72.30, 73.30, 73.723, 75.30, 75.725 • Gastrointestinal tract, interventional procedures, 72.126, 73.126, 75.126 • State of the Art
The use of expandable metallic stents in the treatment of gastric or colonic obstruction is a direct extension of their established utility in treating obstruction within the biliary tree, tracheobronchial tree, vascular system, and esophagus. Within the esophagus, stent placement has been used effectively in the treatment of malignant dysphagia and is now a well-established procedure (1–6). Technical success approaches 100%, and improvement in the dysphagia score ranges from 83% to 100% (1–3). As with the esophagus, metallic stents used within the stomach, duodenum, and colon are intended not to be curative but rather as nonsurgical palliation of the symptoms of obstruction. Unlike treatment in the esophagus, the treatment of gastroduodenal and colorectal obstructions with metallic stents is not yet well established, to our knowledge. Increased familiarity with the available devices, indications, and techniques will be needed before widespread acceptance can be achieved. Innovations in stent design also promise to increase the use of this treatment method in the future.
Unlike plastic tubes or catheters, flexible metallic stents have a high ratio of deployed (expanded) diameter to introduction diameter. That is, they have a relatively small-caliber introduction system that allows safe and atraumatic placement via the mouth or anus; yet when deployed, they expand to a diameter large enough to relieve the obstruction. Some of these devices will even pass through the working channel of a therapeutic endoscope.
STENT TECHNOLOGY
Despite the recent advances in stent technology, the search for the ideal enteral stent continues. Enteral stents should be flexible enough to allow placement but should remain in position once deployed. They should have an internal diameter large enough to relieve obstructive symptoms and restore normal eating and bowel habits. They should have sufficient radial force to expand slowly within areas of fibrosis or neoplasm. They should also prevent obstruction due to tumor ingrowth or reactive hyperplasia. Covered stents have the advantage of resisting tumor ingrowth but tend to be less stable and more rigid, require larger delivery systems, and are more likely to migrate. They are thus more difficult to deploy at distant locations through a tortuous delivery path (7). Uncovered stents are more flexible, and at least one variety can be passed through the working channel of an endoscope. However, when used for long-term palliation of malignant obstruction, they are subject to tumor ingrowth and resultant obstruction.
Until recently, stents used in the stomach and colon were those designed for esophageal, tracheobronchial, or vascular use. Commercially available esophageal stents that have been used in the stomach or colon include the Gianturco Z-stent (Wilson-Cook, Winston-Salem, NC), Wallstent (Boston Scientific, Natick, Mass), Ultraflex nitinol mesh stent (Microinvasive, Natick, Mass), and Esophacoil (IntraTherapeutics, Eden Prairie, Minn) (Fig 1). Tracheobronchial stents that have been used in other anatomic areas include the Gianturco Z tracheobronchial tree stent and the Wallstent tracheobronchial endoprosthesis. A Wallstent enteral endoprosthesis is now available in the United States specifically for the treatment of gastroduodenal and colonic obstructions (Fig 2).
|
|
|
|
The Wallstent is also available in covered and uncovered varieties. There are two silicone-covered esophageal Wallstents. Wallstent I has a dog-bone configuration with an 18-mm luminal diameter and is placed with a 38-F delivery system. Wallstent II has a more gradual flare at each end and is introduced with an 18-F delivery system (Fig 1). The Wallstent esophageal endoprosthesis has a length of uncovered wire at each end to allow tissue ingrowth and stabilization (10). The uncovered Wallstent endoprostheses (tracheobronchial or enteral) are available in a wide range of diameters; the largest expands to 24 mm. The enteral endoprosthesis is deployed from a 10-F delivery system that can fit through a 3.5-mm working channel of an endoscope (Fig 2).
The original Ultraflex nitinol mesh stent is an uncovered stent knitted with elastic alloy wire that slowly expands when deployed at body temperature. It has a lower radial expansile force than the other stents, but its flexibility allows it to be placed across acutely angled stenoses (10,11). The body is 18 mm in diameter and is introduced via a 24-F (outer diameter) sheath (Fig 1). The Ultraflex II is now available in both covered and uncovered varieties and is enclosed within a weave that is released during deployment.
The Esophacoil is an uncovered nitinol coiled stent delivered through a 32-F introduction system.
STOMACH AND DUODENUM
Clinical Factors and Alternative Treatments
Malignant obstruction of the stomach or duodenum causes nausea, vomiting, esophagitis, electrolyte imbalance, poor nutrition, and severe dehydration (12). Causes include primary tumors of the stomach and duodenum, malignant infiltration by neoplasms from adjacent organs (eg, pancreas), and compression by malignant regional lymphadenopathy. Curative resection is not possible in 40% of patients with gastric cancer and 80%–95% of patients with pancreatic cancer (13,14). Palliative gastrojejunostomy with or without gastric resection is commonly performed to relieve obstructive symptoms (13–16). However, in patients with extensive malignant disease and a poor short-term prognosis, surgery may not be possible or desirable (17). Nonsurgical palliative endoluminal techniques such as endoluminal irradiation, laser ablation, and chemical injection have been performed, predominantly in the esophagus. They tend to be useful only in cases of bulky exophytic tumors, often require multiple sessions, and can be accompanied by clinically important complications (18–28).
Nonsurgical palliation can also be accomplished by means of nasogastric intubation or the placement of a gastrostomy or a gastrojejunostomy tube. Prolonged nasogastric intubation is very uncomfortable. A dual-lumen gastrojejunostomy tube will allow gastric decompression and direct jejunal enteral feedings. The gastrojejunostomy tube will not, however, allow the oral intake of solid food (29).
Indications
Expandable metallic stents offer another nonsurgical alternative. They are most useful in poor surgical candidates with malignant obstruction of the gastric outlet. These patients may be medically unable to undergo a surgical procedure or may have a limited life expectancy, which makes the more durable surgical procedure unwarranted. The ability to treat duodenal obstruction secondary to extrinsic compression from pancreatic cancer further expands the spectrum of indications of metallic stent placement for primary nonsurgical palliation. Patients who are able to undergo a more extensive procedure and who have a reasonable life expectancy will continue to benefit from conventional surgical treatment, even when palliative. Metallic stents have also been used in patients with benign gastroduodenal strictures when conventional surgical resection or bypass was not possible or wanted (29–46). In patients with benign disease, coexistent morbid factors involving the cardiopulmonary systems may limit surgical options, which makes the use of metallic stents more attractive. In all cases, intestinal perforation remains the primary contraindication to metallic stent placement.
Technical Considerations
Both covered and uncovered stents have been deployed in the gastroduodenal segment (29–32,42). The primary advantage of covered stents is the prevention of tissue ingrowth (tumor or mucosal hyperplasia) that leads to recurrent obstruction (42,47). This advantage is often outweighed by the stent's increased rigidity and tendency to migrate. Uncovered stents are more flexible and therefore easier to deploy (29–38). The enteral Wallstent can be deployed through the working channel of a therapeutic endoscope (39). Tumor ingrowth through the interstices of the uncovered stent has not been a clinically important disadvantage because of the limited life expectancy of patients with extensive malignancy (29,31,39).
Expandable metallic stents can be placed with fluoroscopic guidance, endoscopic guidance, or a combination of both (29–44). The techniques used will often depend on local political and utilization issues. In programs with a strong and aggressive interventional gastroenterology division, stent insertion with endoscopic or combined endoscopic-fluoroscopic guidance will likely be performed. In areas where an interventional gastroenterologist is not present, is unavailable, or is uninterested or where a direct referral pattern has been established from a surgeon to an interventional radiologist, fluoroscopically guided stent insertion will be performed. Fluoroscopic monitoring is extremely helpful in accurately deploying the stent and ensuring that the lesion is adequately covered.
Fluoroscopic guidance.—All patients should undergo a preprocedural contrast radiologic study of the affected segment of the gut for an anatomic overview. A nasogastric tube is placed and the stomach is decompressed overnight prior to the procedure. The procedure can be performed with the patient under conscious sedation on a conventional fluoroscopic table or within a vascular-interventional suite.
The nasogastric tube is exchanged for a steerable angiographic catheter, which is passed via the nose or mouth (Fig 3). It is often helpful to place a long sheath that extends to the midesophagus to reduce trauma to the oral and hypopharynx during catheter exchanges. Iodinated contrast material is injected with the catheter in the stomach to identify the proximal extent of the stricture. The catheter is manipulated through the obstruction by using standard catheter and guide wire techniques. This is the most technically challenging portion of the procedure, as torque control is often poor when the catheter buckles within a capacious stomach (29,32,38,42,43). Gastric decompression for several days before the procedure can minimize this problem.
|
Once a catheter has been passed beyond the lesion, contrast material is injected again to demonstrate the distal margin of the stenosis. The catheter is manipulated into the proximal jejunum, and an exchange length stiff guide wire is placed. Preplacement balloon dilation is usually not necessary for device deployment. However, gentle balloon dilation may be used to better determine the exact dimensions and location of the proximal and distal aspects of the stenosis when they are not clearly delineated with intraluminal contrast material. These areas can be marked on the fluoroscopic screen or on the patient.
A stent of adequate diameter (at least 16 mm) and length should be chosen to traverse the entire stenosis, with allowance of 1–2 cm extending proximally and distally after expansion. If multiple stents are required, the distal stent should be placed initially, with confirmation that it extends well beyond the lesion. At least 1–2 cm of the stents should overlap to prevent migration.
Postdeployment balloon dilation is usually not necessary, as most stents will self-expand over time. A postinsertion contrast material–enhanced radiographic study is performed through a guiding catheter to assess positioning and to evaluate the need for additional stents. A study with oral contrast material enhancement is performed prior to discharge (Fig 4), and the patient receives instructions relating to eating habits in an attempt to minimize episodes of solid food bolus obstruction (29,30,32,38,42,43). If the procedure was performed via a gastrostomy access, the gastrostomy tube is left in place until there has been adequate adhesion of the stomach to the abdominal wall. At a later time, the gastrostomy tube can be removed if deemed necessary by the patient (29).
|
If the endoscope can be passed through the lesion, a submucosal injection of iodinated contrast material can be performed to mark the distal end of the stenosis (39). An extra-long exchange length stiff guide wire is placed well beyond the lesion. The endoscope is then removed and the stent is deployed over the guide wire. The enteral Wallstent can be deployed through the accessory channel of the endoscope (39). Other larger devices require removal of the scope prior to placing the device over the wire. In these cases, the endoscope can be reinserted adjacent to the device.
By using fluoroscopic markers, the device is deployed while making sure that at least 1–2 cm is distal to the lesion. The proximal location of the stent can be monitored fluoroscopically and can be directly seen with the endoscope within the stomach. Additional stents are deployed if necessary. The final adequacy of stent placement is assessed both endoscopically and fluoroscopically (30,31,34–37,39,44). A final contrast-enhanced radiographic examination is performed following the procedure (Fig 4).
Results and Complications
Soetikno et al (39) reported the results of a prospective study in which enteral Wallstents were placed through an endoscope in 12 patients with malignant gastric outlet obstruction. They used stents ranging from 16 to 22 mm in diameter and 60 to 90 mm in length. Eleven of their patients presented with nausea and vomiting and were unable to take in adequate oral nutrition. One patient with pancreatic cancer and outlet obstruction had recurrence of cholangitis due to food impaction in a biliary Wallstent placed through a choledochoduodenostomy.
All stents were successfully placed. Three patients were treated entirely as outpatients. Following stent placement, six patients were able to eat a regular diet, and three were able to eat a soft pureed diet. Three patients had no relief of symptoms. One of these had multiple distal small-bowel strictures, and two had malpositioned stents. Follow-up ranged from 2 to 40 weeks, during which time nine died of cancer. Complications included three recurrent obstructions—one in a patient with distal migration 2 weeks after implantation treated with repeat stent placement, one in a patient with tumor ingrowth at 21 weeks treated with repeat stent placement, and one in a patient with proximal migration treated supportively. No perforations were reported.
Feretis et al (32) used 22-mm-diameter Wallstents placed with endoscopic-fluoroscopic guidance but not passed through the endoscope in 12 patients. There was complete relief of symptoms in 11 patients. One patient had partial tumor ingrowth; by 6 months after stent insertion, only two patients were still alive.
Binkert et al (31) placed 16-mm-diameter Wallstents with combined endoscopic-fluoroscopic guidance in nine patients, seven with malignant disease and two with chronic ulcer disease. Technical success was achieved in eight (89%) patients. During the follow-up, which ranged from 1 to 52 weeks or until death, there were no signs of stent obstruction. Seven patients had clearly improved quality of life with relief of vomiting and restoration of oral food intake.
Nevitt et al (37) used endoscopic-fluoroscopic guidance to insert four stent designs in eight patients with malignant obstruction. All patients had the distal and proximal aspects of the lesion marked with an intramural injection of contrast material through the endoscope. All stents were deployed successfully. Seven patients had immediate relief of obstructive symptoms. Two had recurrent symptoms due to tumor ingrowth, which was treated with percutaneous gastrostomy in one patient and gastrojejunostomy in the other patient.
De Baere et al (29) reported on fluoroscopically guided insertion of 16-mm-diameter Wallstents without endoscopic assistance in 10 patients. Seven patients had successful peroral insertion of the device. In one, the stricture could not be catheterized via the mouth, and in two, the available stent delivery catheter was not long enough to reach the strictured area via the mouth. These three patients then underwent successful stent placement via percutaneous gastrostomy access. All three patients had a 9-F Foley catheter that was left in place and then removed 10 days after stent insertion. Peroral cannulation required a mean of 17 minutes (range, 3–40 minutes) in the seven patients with successful treatment. In the one patient with treatment failure, oral cannulation was attempted for 45 minutes before reverting to a gastrostomy access. Cannulation then required less than 5 minutes.
Eight of De Baere et al's 10 patients had an immediate benefit from the procedure. In the two without benefit, an examination performed with oral contrast material demonstrated multiple unsuspected small-bowel stenoses that had been previously overlooked because of the severe duodenal stenosis. The stents were beneficial until death (at 25–168 days after implantation) in six patients and remained beneficial in two patients at 55 and 76 days after implantation. The mean hospital stay was 3 days (29). The flexible small-caliber Wallstent delivery system was critical for the successful peroral deployment without endoscopic assistance.
Pinto (38) also reported experience with fluoroscopically guided metallic stent implantation in six patients with malignant gastric outlet obstruction. Five of the six had successful peroral stent placement, and one required gastrostomy access. Wallstents (16 or 20 mm) were used, and all patients were relieved of their presenting symptoms. One had an ascitic leak from the gastrostomy site and died 1 week after the procedure. Three patients had stent obstruction within 2 months of placement. Two patients with stent obstruction were treated with additional stent placement, with relief of symptoms. By 20 weeks, only one patient remained alive. In one patient with recurrent obstruction, endoscopy revealed chronic inflammatory tissue ingrowth without malignant tumor.
Yates et al (30) used a variety of stents (Wallstent, Ultraflex, Esophacoil) to palliatively treat gastric or small-intestinal obstruction in 11 patients, with combined endoscopic-fluoroscopic guidance. Both technical and clinical success was achieved in 10 of the 11 patients (91%). The one failure was due to severe anastomotic angulation and distal luminal obstruction. Five patients had subsequent stent occlusion due to tumor ingrowth at 8
, 3, 4, 11, or 7 weeks after implantation. There was no evidence of perforation, bleeding, or infection.
Summary
Combining the reported cases in 11 series, 91 patients with inoperable gastric outlet obstruction were treated with expandable metallic stents (Table 1) (29–32,34,36–39,48,49). There was immediate clinical benefit in 81 patients (89%). Five of the 10 patients with no clinical benefit were subsequently found to have additional unsuspected stenoses of the small intestine. These synchronous small-intestinal lesions are easy to overlook because of the difficulty in evaluating the small bowel in patients with gastric obstruction. Seventeen patients (21% of the group with an initially successful procedure; 19% of the entire group) had recurrent obstructive symptoms. Stent occlusion secondary to tumor growth and stent migration were the causes of recurrent symptoms in 14 (15%) and three (3%) patients, respectively.
|
Clinical Factors and Alternative Treatments
Malignant tumors of the colon and rectum account for more than 150,000 deaths yearly worldwide (50). Of all patients with colorectal malignancy, 10%–30% have large-intestinal obstruction at presentation (51,52). Furthermore, even in those patients who undergo successful resection, recurrent disease may lead to recurrence of intestinal obstruction.
Acute colonic obstruction is considered a surgical emergency. Failure to treat the obstruction expeditiously predisposes the patient to perforation, metabolic and electrolyte disturbances, intestinal ischemia, and sepsis. The majority of malignant colonic obstructions occur in the descending or sigmoid colon or in the rectum (53,54). Elective resection of a properly cleansed and antibiotically prepared colon is associated with a mortality rate of less than 5%. Emergent colectomy, on the other hand, especially for left-sided malignant obstruction, is associated with much higher mortality (23%) and morbidity (approaching 50%) rates (53,55). The poorer outcomes are largely due to inadequate colonic preparation and other complications associated with emergent surgery. In addition to a higher surgical complication rate, urgent colonic resection without preoperative preparation or adjuvant therapy is associated with a decrease in the overall survival rate (56).
Emergent surgery for colonic obstruction alters the surgical approach to the tumor. The traditional selective approach involves segmental resection of the tumor and drainage of lymph nodes, with primary anastomosis in a single operation. This is ill-advised in patients with poorly prepared colons and markedly dilated proximal intestines. In acutely ill patients, a diverting colostomy may be the best surgical option. In less ill patients, a two-stage operation (Hartmann procedure) is performed. This entails primary tumor resection and end colostomy, with colostomy closure performed as a second, subsequent procedure (51,57). Nearly 25% of patients never undergo the colostomy closure (51). Furthermore, colostomy closure is associated with substantial morbidity and mortality rates, as high as 37% and 7%, respectively (51).
A number of alternative approaches have been advocated to treat patients with malignant colonic obstruction without performing diverting colostomy and multiple operations. Some use percutaneous decompression with irrigation of the bowel. Others advocate a single surgical procedure consisting of segmental resection with on-table lavage followed by primary resection and anastomosis, with an intracolonic bypass tube used to protect the anastomosis (58–61) from breakdown or leakage. Each of these approaches adds complexity to the operation.
Patients with unresectable primary or recurrent malignant colonic obstruction at presentation represent an even greater dilemma. They have limited life expectancy and would clearly benefit from a safe, effective, and nonsurgical alternative to relieve their obstructive symptoms.
The primary goal of a nonsurgical approach for treating acute colonic obstruction is to eliminate the need for emergent surgical treatment of the inadequately prepared colon in an insufficiently stabilized patient (54). Endoscopic laser therapy has been used to maintain luminal patency (62,63). This approach requires a substantial investment in equipment. Because of the risk of perforation and tumor seeding, it is limited to those patients with unresectable tumors. An endoluminal stent with a plastic tube placed transanally via an endoscope has also been used (64–66). Lesions above the rectum require long flexible tubes that have limited internal diameters (65).
Indications
The placement of flexible expandable metallic stents across obstructive lesions of the colon is a new nonsurgical alternative for reestablishing luminal patency (67–80). These devices, already covered, have the large internal diameters needed in the colon. The primary indications for endoluminal metallic stent placement in the colon and rectum are (a) for temporary colonic decompression in patients with an acute potentially resectable malignant colonic obstruction to allow laxative preparation and a single-stage surgical resection and (b) for long-term colonic decompression in patients with unresectable malignant obstruction of the colon or rectum. Patients with diverticulitis and obstruction can also be treated with metallic stents to permit elective colon cleansing prior to resection. It is often difficult to distinguish diverticulitis from tumor as the cause of the obstruction. The presence of tumor does not necessarily need to be established before decompression. Clinical or radiologic evidence of perforation is a contraindication to the procedure. Tumors that are too long or kinked or that are too proximal within the colon are unable to be treated.
Technical Considerations
Stent technology is rapidly evolving, and devices are now being developed specifically for colorectal applications. A variety of stents have been used effectively in the colon, including the enteral Wallstent, Ultraflex stent, Esophacoil, Gianturco-Rosch stent (Cook, Bloomington, Ind), and Memotherm stent (Bard, Covington, Ga) (67–80). Stent placement in distal colonic and rectal lesions can be guided with fluoroscopy or endoscopy alone. Because of the difficulty in accessing more proximal portions of the colon because of sigmoid colon tortuosity, these lesions are best handled with combined endoscopic-fluoroscopic guidance.
A radiologic water-soluble enema (one part 76% contrast material with two parts water) examination is performed to localize and characterize the length and caliber of the obstructive lesion. The enema is also useful for determining the best position in which to place the patient to display the lesion such that the stricture is perpendicular to the beam. This position facilitates visualization of the stricture during stent placement. On the basis of the length of the stricture and its distance from the anus, the appropriate device and delivery system are selected.
Fluoroscopic guidance.—With the patient placed in a good position to visualize the obstructive lesion, an angiographic catheter or guiding catheter is placed into the colon. High-torque, 70–100-cm-long catheters are often needed. The colonic segments are negotiated with the catheter–guide wire combination. Conventional or hydrophilic 0.035- or 0.038-inch wires are used.
Following successful catheterization past the obstructive lesion, contrast material is injected through the diagnostic angiographic catheter to better define the lesion (length, location, optimal position) and to rule out a perforation. The location of the lesion is marked and noted. An exchange length stiff guide wire is advanced well beyond the lesion. When using Wallstents, the tracheobronchial system has a working length of 110 cm, as compared with 240 cm for the enteral system, and may be easier to deploy. A stent with an adequate length and diameter (20–24 mm) is advanced under fluoroscopic guidance and deployed such that the middle of the stent covers the lesion with 1–2 cm of the stent extending beyond both the proximal and distal margins of the lesion (Fig 5). Long lesions may require two stents (69,75–78) (Fig 5).
|
|
|
|
|
|
|
|
|
A water-soluble enema examination is performed immediately after the procedure or the following day to check for patency and leakage (Figs 6d, 7d)(70–73). Patients who are candidates for resection are treated medically, and a bowel-cleaning preparation is instituted for subsequent surgery. Patients with unresectable disease who receive stents only for palliation are placed on a low-residue diet and receive mineral oil to lessen the likelihood of stent obstruction. Follow-up radiographs are obtained in symptomatic patients if needed to check for stent migration, perforation, or colonic obstruction.
Results and Complications
Fluoroscopic guidance, endoscopic guidance, and combined endoscopic-fluoroscopic guidance have all been successfully used to deploy metallic stents within the colon and rectum (67–80). Technical success in deploying the device has ranged from 80% to 100%. Clinical or medical success (relief of obstructive symptoms) has been reported in 80%–92% (Table 2). A higher rate of successful stent placement can be achieved with more distal lesions.
|
Binkert et al (73) reported the results of colonic stent placement in 13 patients with a combined endoscopic-fluoroscopic guidance technique. Eleven of the 13 had lesions in the rectosigmoid region; one patient each had a lesion that was successfully treated with stent placement in the hepatic flexure or descending colon. Uncovered, 16-mm, vascular Wallstents were used with 92% (12 of 13 patients) technical and clinical success rates.
DeGregorio et al (75) reported the results of a prospective multiinstitutional colon stent trial in Europe in which the use of metallic stents for only palliative purposes in patients with acute colorectal obstruction was studied. Twenty-four patients were treated with uncovered stents by using only fluoroscopic guidance. The 22-mm Wallstent was used in 22 patients. One 25-mm Memotherm stent and one covered 18-mm Z-stent were also implanted in one patient each. The technical success rate with fluoroscopic guidance alone was 100% (24 of 24 patients). Twenty-three patients had lesions in the rectosigmoid area, and only one patient had a lesion within the proximal descending colon. Resolution of bowel obstruction was achieved within 24 hours in 96% (23 of 24) of patients. The high success rate can partly be explained by the predominance of distal lesions.
Mainar et al (76) reported the results of the use of metallic stents in the treatment of acute colorectal obstruction prior to surgical resection. In 71 patients, lesions were located in the transverse colon in one patient, descending colon in 22 patients, and rectosigmoid region in 48 patients. All stents were placed within 24 hours of diagnosis of the acute obstruction. Of the 71 patients, 64 (90%) were able to have stents placed in the proper location. In two patients, stents could not be deployed because of the proximal location (transverse colon) in one patient and because of a tortuous rectosigmoid colon in the other. The longer delivery systems now available and the use of endoscopic assistance may have allowed these obstacles to be overcome. In five patients, stents were poorly positioned; three patients required additional stents. Overall, 60 (85%) of the 71 patients had immediate relief of symptoms and of radiographic signs of colonic obstruction within 24 hours of stent deployment. Six other patients were relieved in 24–96 hours. Three (4%) experienced no relief despite apparent proper stent placement and required emergent surgery.
Perforation, stent migration, and restenosis are the major complications encountered with colonic stent placement. Perforation has been reported in 0%–16% of cases (67–79) (Table 2). Baron et al (72) reported colonic perforation in four patients, which was associated with balloon dilation of the stricture at the time of stent placement in three of the patients. Balloon dilation is no longer recommended, and stents are allowed to slowly self-expand. Without balloon dilation, perforation rates may fall below 5%. Perforation should be suspected in patients who complain of abdominal pain during or immediately after the procedure.
Stent migration has been reported to occur in as many as 40% of cases and is usually detected on follow-up radiographs within 1 week of insertion. The displaced stent can be removed endoscopically and another (larger) stent can be placed. Migrations tend to occur with stents too narrow in diameter and/or too short in length for the stricture they are placed in. Stent migration should also decrease substantially in the future with the wider availability of larger-diameter (20–24-mm) stents and flared stents and with increased experience with proper stent positioning.
Choo et al (74) used two covered flexible stents in the colon. One had bulbous ends and was completely covered with polyurethane. This stent migrated in four (50%) of eight patients. The stent was redesigned with the proximal bulbous end left uncovered. There were no cases of stent migration when the redesigned stent was used in subsequent patients.
Restenosis has been reported in up to 25% of cases and is usually due to tumor ingrowth through the open interstices of the stent (67–79). Tumor ingrowth is not a problem in those patients in whom short-term colonic stent placement is performed for preparation prior to definitive surgery. However, for palliative colonic decompression for a longer time, tumor ingrowth with restenosis can be problematic. Covered stents can help address the problem of tumor ingrowth but tend to be less flexible and more difficult to deploy in proximal locations and are associated with higher migration rates (74).
New stents need to be developed to provide long-term colonic decompression by balancing the need to prevent tumor ingrowth and the propensity for stent migration. The redesigned Choo stent (Samsung Biomedical Research Institute, Seoul, Korea) has addressed some of these issues but is not available in the United States. When restenosis due to tumor ingrowth does occur, a new stent can be placed coaxially within the original stent. Restenosis due to tumor overgrowth (extension of tumor above or below the stent) can also be treated with a second stent. It is best prevented by deploying the original stent 2–3 cm above or below the lesion when possible.
Minor complications, including mild abdominal pain and transient rectal bleeding, generally require no treatment. Bare stent wires do produce mucosal injury, but major gastrointestinal bleeding has to our knowledge been uncommon. Fecal impaction can certainly occur and is treated with a cleansing enema. Impactions can be made less likely by using a 20–24-mm-diameter stent and by prescribing a low-residue diet and a mineral oil supplement.
Summary
By combining the reported cases in 12 published articles (Table 2), 234 patients with colonic obstruction were treated with expandable metallic stents (67–70,72–79). The stents were placed with endoscopic, fluoroscopic, and combined endoscopic-fluoroscopic guidance in 25, 136, and 73 patients, respectively. The technical and clinical success rates achieved in this combined group were 93% (217 of 234 patients) and 90% (210 of 234 patients), respectively.
CONCLUSION
Early experience has shown the use of metallic stents within the upper and lower gastrointestinal tracts will provide appropriately selected patients palliation of their obstructive symptoms. In addition, the minimally invasive nature of these procedures will further benefit patients by allowing them to avoid more extensive and invasive surgical procedures. The indications for stent insertion within the gastrointestinal tract will continue to expand with improved stent design and physician ingenuity. Metallic stent placement within the alimentary tract has proved to be beneficial in patients who require palliation of their gastroduodenal or colorectal obstruction and should be a part of the interventionalist's armamentarium.
References
This article has been cited by other articles:
|
|
 |
|
  J. H. Kim, H.-Y. Song, J. H. Shin, E. Choi, T. W. Kim, S. K. Lee, and B. S. Kim Stent Collapse as a Delayed Complication of Placement of a Covered Gastroduodenal Stent Am. J. Roentgenol., June 1, 2007; 188(6): 1495 - 1499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Lopera, A. Brazzini, A. Gonzales, and W. R. Castaneda-Zuniga Gastroduodenal Stent Placement: Current Status RadioGraphics, November 1, 2004; 24(6): 1561 - 1573. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Y. Jeong, J. K. Han, A. Y. Kim, K. H. Lee, J. Y. Lee, J.-W. Kang, T. J. Kim, S. H. Shin, and B. I. Choi Fluoroscopically Guided Placement of a Covered Self-Expandable Metallic Stent for Malignant Antroduodenal Obstructions: Preliminary Results in 18 Patients Am. J. Roentgenol., April 1, 2002; 178(4): 847 - 852. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. H. Baron Expandable Metal Stents for the Treatment of Cancerous Obstruction of the Gastrointestinal Tract N. Engl. J. Med., May 31, 2001; 344(22): 1681 - 1687. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
