Introduction	Choose Top of pageABSTRACTIntroduction <<Subjects and MethodsResultsDiscussionReferencesCITING ARTICLES

Previous reports [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23] suggest that the placement of self-expandable metallic stents is a safe and effective nonsurgical treatment for patients with inoperable gastric and duodenal obstructions. However, recurrent stenosis of the stent because of progressive tumor ingrowth has been a problem because most of the stents used were uncovered. Overall recurrent stenosis rates of 8-46% at an interval of 2-21 weeks have been reported [1, 2, 4, 8,9,10,11, 14, 22].

Covered stents have been used in the esophagus to prevent tumor ingrowth [24] and most likely have contributed to the solution of this problem. Recently, placement of a kind of covered stent was reported to be technically feasible and effective for the palliative treatment of inoperable malignant gastroduodenal obstructions [21, 25]. However, stent migration and unsatisfactory delivery systems for peroral placement have been problematic for covered stents in previous investigations [5, 14, 21, 25].

The purpose of this study is to describe our experience with a series of 18 consecutive patients treated with self-expandable metallic stents for the palliation of malignant antroduodenal obstructions.

Subjects and Methods	Choose Top of pageABSTRACTIntroductionSubjects and Methods <<ResultsDiscussionReferencesCITING ARTICLES

From April 1999 to December 2000, 18 consecutive patients (11 men, seven women; age range, 23-70 years; mean age, 56 years) with severe nausea and vomiting caused by antroduodenal obstruction were referred for fluoroscopically guided placement of covered self-expandable metallic stents. The causes of obstruction were gastric carcinoma in 15 patients, esophageal carcinoma in one, gallbladder carcinoma in one, and cholangiocarcinoma in one. The diagnoses were established by means of endoscopic biopsy (n = 15) and percutaneous needle aspiration biopsy (n = 3). In all patients, the tumors were considered inoperable because of extensive tumor growth (n = 2) and the presence of distant metastases (n = 16). Informed consent was obtained from each patient.

The site and length of the strictures were assessed before stent insertion by means of a barium study and CT. The stricture site was gastric antrum to pylorus in 11 patients, gastric antrum to the second portion of the duodenum in five patients, gastric body to the pylorus in one patient, and the second portion of the duodenum in one patient.

Topical anesthesia of the pharynx was achieved routinely before the procedure using an aerosol spray of lidocaine (Xylocaine, 10% spray [10 mg/dose]; Astra Pharmaceuticals, Södertälje, Sweden) and lidocaine hydrochloride oral topical solution (Lidocaine Viscous, 2%; Taejoon Pharmaceuticals, Seoul, Korea). Sedatives were not used in any patient for aspiration during the procedure. A nasogastric tube was placed before the procedure to empty the stomach in patients with severe gastric distention.

The covered stents used were commercially available Niti-S stents (TaeWoong Medical, Seoul, Korea) (Fig. 1) (n = 20 in 17 patients) and Song stents (Doosung Medi-tech, Seoul, Korea) (n = 2 in one patient). The stents were 15 or 18 mm in diameter when fully expanded and 60-170 mm long (mean, 123.1 mm), and both ends of the stent were flared up to 20 mm. A stent at least 20 mm longer than the stricture was selected for placement so that its proximal and distal parts would rest above and below the stricture, respectively. The stent was mounted on a delivery system composed of a guiding tip, an introducing tube, and a pushing catheter 6 mm in diameter and 150 cm in length.

View larger version (92K)

Fig. 1. —Photographs shows polyurethane-covered Niti-S (TaeWoong Medical, Seoul, Korea) stent before deployment with delivery system (top) and after deployment (bottom).

With fluoroscopic guidance, we positioned and deployed stents initially using a peroral route, as described in previous reports [21, 25, 26], which involved using a 100-cm 5-French Berenstein-type vascular catheter (Bard, Tewskbury, MA); a 260-cm, 0.035-inch in diameter exchange guidewire (Radiofocus M; Terumo, Tokyo, Japan); and a 260-cm, 0.035-inch Amplatz Super stiff guidewire (Boston Scientific/Medi-Tech, Natick, MA). In two patients whose peroral placements failed, stent access to the stomach was achieved with percutaneous gastrostomy using standard techniques [14, 27]. Balloon catheters (diameter, 15-20 mm; Ultra-thin diamond balloon dilatation catheter; Boston Scientific/Medi-Tech) were used in three patients to allow the delivery system to pass the stricture. In one patient, a balloon catheter was used immediately after the stent insertion to reposition the stent more precisely and to facilitate the stent expansion.

All patients underwent a barium study 1 day after stent placement to verify the position and patency of the stents. Technical success was defined as precise placement of the stent in the targeted lesion site, as well as its patency to relieve the obstruction, as shown by the contrast barium study. Further follow-up in each patient was based on monthly clinical examinations in the outpatient clinic or in telephone interviews. A follow-up barium study or endoscopy was performed only in patients with recurrent symptoms.

Results	Choose Top of pageABSTRACTIntroductionSubjects and MethodsResults <<DiscussionReferencesCITING ARTICLES

Stent placement was technically successful and well tolerated in all patients, and no major procedural complications occurred (Fig. 2A,2B). Two patients complained about mild abdominal pain immediately after the stent placement, and their pain was controlled with 50 mg/day IV of pethidine hydrochloride (Demerol; Keukdong Pharmaceuticals, Seoul, Korea) for 2 days. In all patients, the negotiation of the guidewire through the obstruction was successful, and the delivery system passed the stricture segment with (n = 3) or without (n = 15) the balloon dilatation procedure.

View larger version (196K)

Fig. 2A. —47-year-old man with gastric carcinoma involving duodenum. Anteroposterior radiograph obtained during stent placement shows diffuse antroduodenal narrowing (arrows). Contrast media was injected through 5-French catheter. Percutaneous biliary drainage catheter is also noted.

View larger version (154K)

Fig. 2B. —47-year-old man with gastric carcinoma involving duodenum. Anteroposterior upper gastrointestinal radiograph obtained 1 day later shows good passage of contrast media.

Symptoms improved immediately after the procedure in all patients except one. This patient had another stenosis at the proximal jejunum that had been concealed before stent placement. She required IV nutrition until her death. The other 17 patients were able to take liquids or solid foods after the stent placement, and 11 of them subsequently were able to eat any type of food. There was no clinical or radiologic evidence of perforation of hollow viscus during the procedure or follow-up period.

Stent migration occurred in three patients (16.7%). In two patients, the stents migrated downward at 1 and 41 days, respectively, after the procedure. One of the stents came out of the anus uneventfully 13 days after placement, and a second covered stent was placed at the stricture site. In the other patient with downward migration, a second covered stent was coaxially inserted in a telescoped manner overlapping the stent that had previously migrated. In the third patient (Fig. 3A,3B,3C,3D), the stent migrated upward into the gastric lumen 14 days after the procedure. Endoscopic removal of the stent failed, and it remained inside the stomach until the patient's death without causing symptoms. A second stent was inserted at the stricture site. Neither further migration nor obstructive symptoms recurred in these three patients during the remainder of the follow-up period.

View larger version (164K)

Fig. 3A. —63-year-old woman with gallbladder carcinoma invading antrum and duodenum. Anteroposterior upper gastrointestinal radiograph obtained before stent placement shows total obstruction of gastric outlet with food materials in distended stomach.

View larger version (184K)

Fig. 3B. —63-year-old woman with gallbladder carcinoma invading antrum and duodenum. Anteroposterior upper gastrointestinal radiograph shows well-positioned stent 1 day after stent placement.

View larger version (206K)

Fig. 3C. —63-year-old woman with gallbladder carcinoma invading antrum and duodenum. Right anterior oblique upper gastrointestinal radiograph shows good barium flow immediately after placement of second covered stent (2 weeks after initial procedure). Initial covered stent migrated proximally into gastric lumen (arrow). Endoscopic removal of stent failed and remained inside patient's stomach until her death without causing symptoms.

View larger version (179K)

Fig. 3D. —63-year-old woman with gallbladder carcinoma invading antrum and duodenum. Follow-up conventional radiograph shows that patent lumen of second stent is established.

In two patients, the stents (two Niti-S stents) were collapsed during the follow-up imaging studies 30 and 64 weeks after stent placement, causing recurrent symptoms (Fig. 4A,4B,4C). A second covered stent was coaxially placed in one of these patients, which resulted in a positive clinical outcome; and the mechanical failure did not recur. Another patient refused additional management for a collapsed stent, and she died as a result of sepsis 8 weeks later.

View larger version (170K)

Fig. 4A. —63-year-old woman with advanced gastric carcinoma involving antrum and pylorus. Left anterior oblique upper gastrointestinal radiograph obtained before stent placement shows nearly total obstruction of antrum of stomach. Although guidewire and catheter could pass beyond stricture, stent delivery device could not follow peroral route because of markedly distended J-shaped stomach.

View larger version (176K)

Fig. 4B. —63-year-old woman with advanced gastric carcinoma involving antrum and pylorus. Right anterior oblique upper gastrointestinal radiograph shows good barium passage 1 day after stent placement. Stent was placed through percutaneous gastrostomy made in gastric body.

View larger version (216K)

Fig. 4C. —63-year-old woman with advanced gastric carcinoma involving antrum and pylorus. Anteroposterior upper gastrointestinal radiograph obtained 64 weeks later shows collapsed stent with recurrent obstruction.

During the follow-up period (range, 2-73 weeks; mean, 12 weeks), 16 of the 18 patients died (range, 2-73 weeks; mean, 12.1 weeks) after stent placement owing to sepsis or diffuse metastasis of the underlying cancer, and two patients were alive without obstructive symptoms 14 and 31 weeks after stent placement.

Discussion	Choose Top of pageABSTRACTIntroductionSubjects and MethodsResultsDiscussion <<ReferencesCITING ARTICLES

Malignant antroduodenal obstruction is among the many serious conditions that can lead to nutritional deficiency and dehydration [22]. Although palliative bypass surgery is an available option in such patients, the morbidity and mortality in performing palliative surgery are high, and control of symptoms is achieved in about only half of the patients treated, especially when curative resections are not possible [28,29,30,31,32,33,34,35]. Different types of uncovered metallic stents have been reported to provide alternative treatment with minimal morbidity in these patients [1,2,3,4,5,6,7,8,9,10,11,12,13]. Recent studies reported successful placements of covered self-expandable metallic stents for malignant gastroduodenal obstructions [21, 25]. Covered stents obviously have the theoretic advantage of a lower incidence of tumor ingrowth compared with uncovered stents.

The requirement of better designed stents and delivery systems for peroral placement of covered stents has been emphasized because the systems used in the esophagus are too large and rigid to be used in the gastric outlet or duodenum [1, 21, 25]. In our series, peroral placement of standard covered stents was technically successful in most of the patients (88.9%), and gastrostomy was required in only two patients. In those patients with a markedly distended stomach or J-shaped stomach, peroral placement was difficult because stent systems kinked in the stomach even with extra-stiff guidewire. Using gastrostomy as a route of stent delivery, we found that the passage of the stenotic segment was easier, and patients tolerated the procedure even better than the peroral approach. After 2 weeks, the gastrostomy tube was removed without complications in both patients.

Despite their ability to prevent tumor ingrowth, covered stents have been reported to migrate more often in the esophagus than uncovered stents [5, 14, 21, 26]. Moreover, in the gastroduodenal obstructions, recent studies [21, 25] have indicated that a high migration rate (27.8% and 20.8%, respectively) was the major problem in their results. Although we agree that further investigation is necessary to modify the covered stent to prevent migration, the lower migration rate (16.7%) in this study is encouraging compared with those of previous reports [21, 25].

Interestingly, collapse of the stents was noted in two patients in this study, which probably resulted from the centripetal growth of the tumor or the fatigue fracture of the stent wires. This kind of mechanical failure is different from the previously reported stent disruption of modified Z stents (Choo stent; Mi Tech, Pyungtaek, Korea) associated with the tearing of the polyurethane cover [21]. To our knowledge, this mechanical failure has not been reported, at least with covered stents placed in the gastrointestinal tract. Because of continuous peristaltic movement around the stent, together with the corrosive nature of gastric secretion, there is a chance of mechanical failure in long-term survivors (> 30 weeks in our study). Improvements in wire material and stent design are required.

In conclusion, our results indicate that the placement of covered self-expandable metallic stents is a safe, technically feasible, and effective treatment palliating inoperable malignant antroduodenal obstructions. The incidence of migration as a complication of this procedure was lower than those reported in the literature.