INTRODUCTION — Self-expandable metal stents are a nonsurgical alternative for the palliation of luminal gastrointestinal neoplasms, particularly in the esophagus and biliary tract. They also have a role in the palliation of obstruction in other segments of the gastrointestinal tract such as the stomach, proximal small bowel, and colon (enteral stents).
In the Western world, malignant gastric outlet obstruction occurs most commonly from obstruction of the duodenum by pancreatic cancer. Other causes at the level of the duodenum include cholangiocarcinoma and metastatic disease. Malignant gastric obstruction at the level of the stomach due to gastric cancer (primary or postsurgical recurrence) is seen more commonly in Asian countries [1].
Gastroduodenal obstruction causes significant morbidity, including nausea, intractable vomiting, esophagitis, electrolyte imbalances, poor nutrition, and severe dehydration. In many cases, stenting can provide durable symptom relief without the morbidity associated with palliative surgery, and it may provide a better quality of life compared with other alternatives.
This topic will review the role of enteral stents for the palliation of malignant gastroduodenal obstruction. An overview of gastric outlet obstruction, as well as topics on stenting of the esophagus, biliary tract, and colon are discussed separately.
●(See "Gastric outlet obstruction in adults".)
●(See "Endoscopic stenting for palliation of malignant esophageal obstruction".)
●(See "Endoscopic stenting for malignant biliary obstruction".)
●(See "Enteral stents for the management of malignant colorectal obstruction".)
The management of patients with advanced pancreatic or gastric cancer is discussed separately:
●(See "Supportive care of the patient with locally advanced or metastatic exocrine pancreatic cancer".)
●(See "Local palliation for advanced gastric cancer".)
INDICATIONS — The goal of enteral stent placement for malignant gastroduodenal obstruction is to restore luminal patency in symptomatic patients with an unresectable malignancy or recurrence of malignancy at an anastomotic site. Patients being considered for stent placement should have a short life expectancy (less than two to six months) [2,3]. One method for estimating survival uses the World Health Organization (WHO) performance status:
●WHO 0: Fully active
●WHO 1: Cannot carry out heavy physical work
●WHO 2: Up and about more than 50 percent of the day
●WHO 3: Up and about less than 50 percent of the day
●WHO 4: Bed or chair bound all day
In a study of 105 patients with malignant gastroduodenal obstruction, patients with a performance status of 3 or 4 had a three-month survival rate of 26 percent, compared with 60 percent for those with a performance status of 0 to 2 [4]. Surgical palliation with a gastrojejunal anastomosis should be considered for patients with longer life expectancies (predicted by a WHO performance status of 0 to 2) because it offers better long-term relief compared with stent placement.
The goals of stent placement are to provide relief from obstructive symptoms, to allow the patient to resume a normal diet, and to improve the patient's quality of life. As such, stent placement is not indicated for patients with a gastric outlet obstruction who are asymptomatic and are tolerating a normal diet. Other contraindications to stent placement are the presence of multiple sites of small bowel obstruction (since stenting the proximal stricture is unlikely to provide symptom relief) and the presence of free perforation.
Patients with peritoneal carcinomatosis are at high risk for incomplete palliation following stent placement given the risk of multiple sites of luminal obstruction. Despite this anecdotal experience, some data suggest that stents can work in such patients. In a retrospective study of 215 patients undergoing stent placement for malignant gastric outlet obstruction, 116 patients (60 percent) had carcinomatosis and 76 (40 percent) did not [5]. Outcomes were similar for those with and without carcinomatosis with regard to clinical success (81 versus 84 percent, respectively), need for reintervention (18 versus 27 percent, respectively), and major complications (4 percent for both groups).
PRE-PROCEDURE EVALUATION — Obtaining a radiographic contrast study (computed tomography [CT] scan or magnetic resonance imaging with oral contrast or upper gastrointestinal series) to assess the anatomy, length of the stricture, and degree of obstruction may be helpful, especially in cases where it is unclear if the obstruction is clinically significant [6]. However, the majority of patients will have had imaging (such as a CT scan) showing the extent of the lesion, thereby mitigating the need for additional imaging.
Prior to gastroduodenal stent placement, one should evaluate the status of the biliary tree. Coexistent biliary obstruction is commonly present since it typically develops earlier than gastric outlet obstruction (eg, patients with pancreatic cancer) [7]. If there is known or impending (ie, radiographic only) biliary obstruction, a self-expandable metal biliary stent should be placed prior to placing the duodenal stent because it is difficult to obtain biliary tree access once a duodenal stent has been placed across the papilla [8,9]. A percutaneous transhepatic or endoscopic ultrasound-guided approach may be required to treat a biliary obstruction that develops after duodenal stent placement [10]. (See "Endoscopic stenting for malignant biliary obstruction".)
EQUIPMENT
Types of stents — Duodenal self-expandable metal stents (SEMS) are used for the palliation of malignant gastroduodenal obstruction. SEMS may be uncovered (meshwork is bare wire) or covered (meshwork is covered to decrease tissue growth into the stent). All duodenal SEMS function similarly.
Duodenal stents available in the United States are all uncovered, have delivery systems of small diameter that can be passed through the working channel of a therapeutic endoscope, are long enough to be passed through a colonoscope, and are reconstrainable up to a certain point during deployment [11,12]:
●The WallFlex Duodenal Stent (Boston Scientific) is composed of nitinol, which is more flexible than the alloy metals used in legacy Wallstent. The WallFlex stent has a diameter of 22 mm through its body, with a 27 mm proximal flare. It is available in lengths of 6, 9, and 12 cm.
●The Hanarostent LowAx Duodenal Stent (Olympus America) is also made from nitinol using a unique hook-and-cross stent weave which aims to optimize the axial and radial force, and allow conformation to the anatomy. It is available in diameters of 22 mm and in lengths of 6, 9, and 12 cm.
●The Evolution Duodenal Stent (Cook Medical) is composed of nitinol and has lengths of 6, 9, and 12 cm. It is unique in that it has both proximal and distal flanges of 27 mm and has a midbody diameter of 22 mm.
Additional stents are available outside the United States, some of which are covered. (See 'Uncovered versus covered stents' below.)
Uncovered versus covered stents — We use uncovered self-expandable metal stents (SEMS) for the treatment of malignant gastroduodenal obstruction. Covered SEMS that are available outside the United States are less flexible and more prone to migration than uncovered SEMS, but covered SEMS offer the advantage of less tumor ingrowth [13,14]. When tumor ingrowth occurs, luminal obstruction may recur in patients who survive for more than six months after stent placement. However, stent occlusion can be managed endoscopically with methods such as tumor ablation and placement of another SEMS within the occluded SEMS [15,16]. Uncovered SEMS also allow for bile flow through the stent interstices in patients with previously placed biliary stents. (See 'Types of stents' above.)
Data have suggested that uncovered and covered SEMS are both effective for relieving malignant gastroduodenal obstruction but that there are differences in complication rates. In a meta-analysis of five trials including 443 patients with malignant gastric outlet obstruction, rates of resuming an oral diet were not significantly different for patients with covered SEMS compared with uncovered SEMS (92 versus 91 percent) [17]. Patients with covered SEMS had lower rates of stent occlusion compared with uncovered SEMS (4 versus 25 percent), while rates of stent migration were higher with covered SEMS (12 versus 3 percent). In a subsequent meta-analysis of five trials and eight observational studies including over 1500 patients with malignant obstruction, the perforation rate was higher with covered SEMS (2.2 versus 0.6 percent); however, there were no significant differences in rates of other adverse events including bleeding, pancreatitis, or cholangitis [18].
PLACEMENT
Personnel — Duodenal stents should be placed by experienced therapeutic endoscopists in a procedure room equipped with fluoroscopy [19]. It is imperative to have a procedural assistant who is familiar with complex, therapeutic endoscopic procedures such as endoscopic retrograde cholangiopancreatography (ERCP) and self-expandable metal biliary stent placement.
The duration of the procedure is highly variable and depends upon the degree of difficulty encountered when traversing or accessing the stricture. At least 60 minutes should be allotted for the procedure after adequate sedation has been achieved.
Procedural sedation and patient preparation — Moderate procedural sedation with an opiate (eg, meperidine or fentanyl) and benzodiazepine (eg, midazolam) is usually sufficient. The majority of patients can be discharged following the procedure without the need for hospitalization. (See "Gastrointestinal endoscopy in adults: Procedural sedation administered by endoscopists".)
In patients who are not endotracheally intubated, the patient should be placed in the left lateral decubitus or prone position to minimize the risk of aspiration due to retained gastric contents. Similarly, the supine position should be avoided unless the patient is endotracheally intubated and/or the gastric contents have been completely evacuated because of the high risk of aspiration. A prone position allows for a better anatomic view under fluoroscopy compared with the left lateral position.
Patients should take nothing by mouth for at least six hours prior to the procedure, and should not eat solid foods for at least eight hours prior to the procedure. However, since these patients have gastric outlet obstruction, gastric clearance times are prolonged. We initiate a clear liquid diet one or two days before the procedure to allow emptying of retained solids. Some authors suggest insertion of a nasogastric tube 24 to 48 hours prior to the procedure to decompress the stomach and remove gastric contents in the presence of complete gastric outlet obstruction [20]. Decompression of the stomach is important as placement of a duodenal stent can be significantly more challenging in the setting of a very dilated stomach. The nasogastric tube can be removed immediately prior to endoscopy and should not be required after stent placement.
Pre-deployment evaluation — As the endoscope is passed to the site of obstruction, a gentle attempt should be made to advance it through the obstruction. Excessive force or aggressive dilation of the stricture increases the risk of perforation and should be avoided in most cases, since stent placement can usually be performed without the endoscope traversing the obstruction. If the endoscope can traverse the obstruction, it should be used to measure the length of the obstruction, so that an appropriate stent can be chosen. If the endoscope is unable to pass through the lesion, the length of the obstruction can be determined fluoroscopically by simple contrast injection. Alternatively, an occlusion balloon catheter can be utilized. Using this accessory, the endoscope is positioned immediately proximal to the obstruction and the catheter is advanced beyond the obstruction. The balloon is inflated and retracted to the distal extent of the obstruction. The length of the obstruction can then be measured with fluoroscopy, or by measuring the length of catheter after retracting the deflated balloon. In cases where the goal is to reach the papilla for the treatment of concomitant biliary obstruction, the duodenal obstruction may need to be traversed. In such cases, dilation may be required [9].
Stent deployment — If the endoscope passes easily through the lesion, a 0.035 inch guidewire with a floppy tip is placed through the endoscope channel and passed at least 20 cm distal to the obstruction and coiled in the small bowel. If the diameter of the endoscope channel is large enough to allow the stent delivery system to pass (ie, typically 3.7 mm or greater), the stent is placed over the guidewire and then advanced through the endoscope. It is then placed across the obstruction under endoscopic and fluoroscopic guidance.
If the stent delivery system is too large to pass through the endoscope channel (eg, non-dedicated duodenal stents such as esophageal stents), a much stiffer 0.035 or 0.038 inch guidewire is placed and the endoscope is withdrawn, leaving the guidewire in place. The stent is then loaded onto the guidewire and advanced fluoroscopically to the obstruction. The endoscope can then be reinserted alongside the stent delivery system to allow direct endoscopic visualization during deployment.
If the endoscope cannot be passed easily through the lesion, a hydrophilic biliary guidewire preloaded through standard biliary accessories can be used to traverse the stricture, as is done during ERCP. Once the wire has passed through the stricture as demonstrated fluoroscopically (which can be recognized by the anatomically correct position of the wire passing into an air-filled distal bowel loop and/or injection of contrast), the catheter is advanced over the guidewire through the obstructing lesion. Water-soluble radiographic contrast is then injected to confirm proper position. The pre-deployed stent is passed over the guidewire and the procedure is continued as described above.
Post-deployment evaluation — Once the stent is fully deployed, the ends of the stent should be carefully inspected fluoroscopically. If either end of the stent is not flared or expanded to produce a "waist" within the obstructing lesion, the endoscopist should be suspicious that the chosen stent may be too short to completely traverse the length of the stricture. At this point, contrast can be injected into the stent to assess complete patency. If needed, a second (and rarely a third) overlapping stent can be placed to adequately treat the stricture. In general, if the endoscope does not need to be passed through the stent immediately after deployment (which may be required if the papilla needs to be accessed for management of the biliary tree), passage of the endoscope through the stent should be avoided. Passage of the endoscope through the stent before the stent has fully expanded may lead to stent displacement.
POSTPROCEDURE CARE — Complete stent expansion usually occurs in 24 to 48 hours, although with very tight strictures, expansion may take longer or the stent may not fully expand. Patients are allowed to ingest liquids following the procedure and to then cautiously advance their diet to a low-residue diet (table 1) [7]. However, there are limited data for guiding the optimal dietary approach and some patients can tolerate a normal diet [7]. Patients with complete, prolonged obstruction may have a delay in return of normal gastric motility. In addition, factors such as neural involvement by tumor, prior gastric resection, and placement of the proximal end of the stent within the antrum rather than entirely in the duodenum may determine which diet can be tolerated.
For patients with persistent symptoms or recurrence of symptoms after initial resolution, imaging (computed tomography scan with oral contrast or upper gastrointestinal series with small bowel examination) is generally performed to assess for issues such as inadequate relief of primary obstruction, tumor ingrowth into the stent lumen, or obstruction distal to stent placement.
EFFICACY — Stents can be placed successfully in over 90 percent of patients, with clinical success rates of typically 80 percent or higher [21-26]. Technical failure is usually due to the inability to pass a guidewire through the stricture, anatomic difficulties such as excessive looping within the dilated stomach, or complicated postsurgical anatomy [6].
Published experience with enteral stenting for gastroduodenal obstruction is derived mostly from observational studies [21,23-25,27-37], although small randomized trials have compared stent placement with surgical bypass for palliation of malignant gastric outlet obstruction [2,38,39] (see "Supportive care of the patient with locally advanced or metastatic exocrine pancreatic cancer"):
●In a review of two trials of endoscopic stenting versus palliative gastrojejunostomy, six comparative studies, and 36 retrospective series, there were no statistically significant differences between the two procedures in terms of efficacy or complications [22]. However, stenting was associated with a trend toward shorter hospital stay, a higher clinical success rate, and faster relief of obstructive symptoms [22]. Patients who received stents did require reintervention more frequently than did surgically-treated patients.
●In a subsequent trial, 39 patients with malignant gastric outlet obstruction were assigned to either gastrojejunostomy (18 patients) or enteral stent placement (21 patients) [2]. Food intake improved more rapidly in the stent group compared with the surgery group (median five versus eight days), but long-term relief was worse in the stent group (median duration of relief 50 versus 72 days). Major complications were more common in the stent group (six complications in four patients compared with no major complications in the surgical group). There were no differences in the groups with regard to survival or health-related quality of life scores.
●In another randomized trial, 18 patients were assigned to endoscopic stent placement or surgical bypass [38]. Endoscopic stenting was superior to surgery with regard to median operative time (40 versus 93 min), mean time to restoration of oral intake (2.1 versus 6.3 days), and median length of hospital stay (3.1 versus 10 days). There were no significant differences between the groups with regard to morbidity, mortality, gastric emptying, or clinical outcomes at three-month follow-up.
●A database study that compared 425 stenting procedures with 339 surgical bypasses for malignant gastric outlet obstruction found that the median length of stay was shorter for stenting procedures than for surgery (8 versus 16 days), and the median cost was lower [40].
Despite initial success, 15 to 40 percent of patients require reintervention for recurrent symptoms or biliary obstruction following stent placement, whereas reintervention rates are generally lower in patients who undergo gastrojejunostomy [2,22,36,37]. Furthermore, some patients may not improve after successful stent placement. This may occur because of unidentified distal sites of malignant obstruction from diffuse peritoneal carcinomatosis with bowel encasement or omental tumor deposits, or gastric outlet obstruction from neural (celiac axis) tumor involvement [6,7,10].
ADVERSE EVENTS — Several adverse events can occur during or after duodenal stent placement [41,42] (see 'Uncovered versus covered stents' above):
●Intraprocedural adverse events include those related to sedation, pulmonary aspiration, stent malposition, perforation, and bleeding
●Late adverse events include distal stent migration, stent occlusion, bleeding, perforation, fistula formation, and occlusion of biliary stents.
For patients with self-expandable metal stents (SEMS), reported rates of serious adverse events (eg, perforation) have been low. In a systematic review of five trials and eight observational studies including 874 patients with malignant gastric outlet obstruction treated with uncovered SEMS, rates of perforation or bleeding were low (0.6 and 1.1 percent, respectively) [18]. In an earlier systematic review with 606 patients, there was no procedure-related mortality [43].
Patients who have a self-expanding metal stent for biliary obstruction and who subsequently undergo duodenal stent placement are at increased risk of biliary stent dysfunction. In a series of 410 patients with biliary stents, 33 underwent duodenal stent placement [42]. Biliary stent dysfunction developed in 17 (52 percent) and duodenal stent placement was a risk factor for biliary stent dysfunction on multivariable analysis (HR 2.0, 95% CI 1.2-3.5). Biliary stent dysfunction occurred in patients with duodenal stents after a median of 64 days. By comparison, the median time to stent dysfunction in patients who did not undergo duodenal stent placement was 170 days.
There have been no data concerning the safety of gastroduodenal stenting in patients who have undergone or are receiving radiation therapy, a setting in which stenting may be riskier [6].
OTHER THERAPEUTIC OPTIONS — Alternatives to enteral stent placement for alleviation of symptoms include palliative surgical resection, surgical bypass (gastrojejunostomy) and radiation therapy, and these options are discussed separately. (See "Local palliation for advanced gastric cancer", section on 'Therapeutic options for local palliation'.)
An endoscopic ultrasound-guided gastrojejunostomy has also been described [44]. This procedural technique allows endosonographic placement of a lumen-apposing metal stent (ie, Axios, Boston Scientific) to create a gastrojejunal anastomosis, which bypasses the gastroduodenal obstruction. In a study of 26 patients with gastric outlet obstruction (from both benign and malignant causes), technical success was achieved in 24 patients (92 percent) and clinical success was achieved in 22 patients (85 percent). Adverse events occurred in three patients (12 percent). While the data are promising, further studies comparing this approach to luminal stent placement are warranted prior to widespread adoption.
SUMMARY AND RECOMMENDATIONS
●Malignant gastroduodenal obstruction can cause significant morbidity, including nausea, intractable vomiting, esophagitis, electrolyte imbalances, poor nutrition, and severe dehydration. Enteral stent placement can provide palliation without the morbidity associated with palliative surgery and may provide a better quality of life compared with other alternatives. (See 'Introduction' above.)
●Candidates for stent placement include patients who have a short life expectancy (ie, less than two to six months). The goals of stent placement are to provide relief from obstructive symptoms, to allow the patient to resume a normal diet, and to improve the patient's quality of life. As such, stent placement is not indicated for patients who are asymptomatic and are tolerating a normal diet. (See 'Indications' above.)
●For patients with malignant gastroduodenal obstruction who are managed with endoscopic stent placement, we suggest uncovered self-expandable metal stents (SEMS) rather than covered SEMS (Grade 2C), given data suggesting similar efficacy at relieving luminal obstruction, but lower rates of stent migration and perforation. For patients with uncovered SEMS who develop tumor ingrowth, luminal obstruction can often be managed endoscopically. (See 'Equipment' above.)
●Enteral stents for the treatment of malignant gastroduodenal obstruction can be placed successfully in over 90 percent of patients. Enteral stents for palliation are associated with less morbidity, procedure-related mortality, and cost compared with surgical palliation. However, some patients require reintervention following stent placement for recurrent symptoms or biliary obstruction. (See 'Efficacy' above.)
●Adverse events can occur during or after stent placement. Intraprocedural adverse events include complications related to sedation, pulmonary aspiration, stent malposition, perforation, and bleeding. Postprocedure adverse events include distal stent migration, stent occlusion, bleeding, perforation, and fistula formation. (See 'Adverse events' above.)
