We retrospectively reviewed our hospital’s database to identify all patients with malignant duodenal obstruction who underwent endoscopic SEMS placement at Seoul National University Boramae Hospital between January 2003 and June 2013. All the patients had symptomatic obstruction characterized by nausea, vomiting, reduced oral intake, and weight loss. In patients with suspected biliary obstruction-based on liver function tests and computed tomography (CT) scan-endoscopic biliary stenting using a metallic stent was always attempted prior to duodenal stenting.

All patients underwent CT to determine the extent of the tumor and to evaluate the site, degree, and length of the obstructive lesion. The histopathological diagnosis was confirmed by endoscopic biopsy that was performed before or during stent placement.

The inclusion criteria were acute malignant gastric outlet obstruction, diagnosed based on the clinical obstructive symptoms and a radiologic examination. The exclusion criteria included asymptomatic patients with duodenal obstruction, evidence of perforation or peritonitis, concomitant small bowel obstruction, or duodenal obstruction caused by benign strictures.

The present study was performed in accordance with the guidelines of our institutional review board, which approved this study.

The stent length was selected by allowing for an additional 2-4 cm to be exposed distally and proximally to the obstructive lesion. The 3 stent types used in our study were as follows: (1) the Niti-S D-type pyloric/duodenal stent (uncovered SEMS; Taewoong Medical, Seoul, Korea); (2) the Niti-S Head-type pyloric/duodenal stent (covered SEMS; Taewoong Medical); and (3) the newly developed Niti-S Comvi-type pyloric/duodenal stent (covered SEMS; Taewoong Medical). The uncovered Niti-S D-type stent was made from a mesh of a single strand of nitinol wire formed into a cylindrical shape without flared ends. The covered Niti-S Head-type stent had a silicone covering with uncovered flared ends. The covered Niti-S Comvi pyloric/duodenal stent had a polytetrafluoroethylene (PTFE) covering with uncovered ends. The inserted stents had diameters of 18 mm, 20 mm, 22 mm, and 24 mm and lengths of 60 mm, 80 mm, 100 mm, and 120 mm.

Two expert endoscopists (Jeong JB, Kim JW) inserted all the stents using an endoscopic approach. For the endoscopic stent placement, the endoscope (GIF-2T240 and CF-260, Olympus, Co., Tokyo, Japan) was carefully inserted into the lesion site. One or two clippings were subsequently performed > 2 cm proximal to the lesion. A 0.035-inch guide wire (Trace metro, Cook, United States) was passed across the obstruction and into the distal part of the obstruction. After a 5F biliary catheter was passed through the guide wire across the obstruction, the length of the obstructive lesion was measured by the injection of water-soluble contrast dye (Gastrograffin) through a 5F biliary catheter. A stent delivery system was advanced over the guide wire through the working channel of the endoscope and was inserted in the obstruction by fluoroscopic guidance. The stent was deployed at the stricture site while pulling back the outer sheath, under combined fluoroscopic and endoscopic guidance.

After stent placement, plain radiographs of the abdomen was obtained at 24 h, 48 h, and 72 h, and we evaluated the position of the stent and the relief of the gastric outlet obstruction.

Technical success was defined as successful stent placement across the obstructive lesion. Clinical success was defined as an increase based on the gastric outlet obstruction scoring system (GOOSS), with a score of > 1 point and/or reduction of poor oral intake or vomiting 1 wk after SEMS placement. GOOSS assigns a point score depending on the patient’s level of oral intake (none, 0; liquids only, 1; soft solids, 2; low-residue or full diet, 3)[20]. Complications after stent insertion were defined as adverse events that induced the recurrence of obstructive symptoms, such as stent occlusion, stent migration, perforation, or biliary obstruction. Stent patency was defined by the period between the initial stent placement and loss of stent function as a result of the complications.

In patients with palliative stent placement, clinical follow-up or telephone interviews regarding the recurrence of obstructive symptoms were conducted at an interval of 1-3 mo.

The duration of stent patency was defined as the time from stent placement to stent-related complications. When no stent-related complication occurred, the duration of stent patency was considered equal to the survival duration.

Patients were divided into 2 groups according to stent type (uncovered and covered). Differences between the groups were analyzed using the Mann-Whitney U test for continuous variables and the χ² or Fisher exact test for categorical variables. Wilcoxon signed-rank tests were used to assess improvements in the GOOSS scores. Overall stent patency and patient survival were estimated using the Kaplan-Meier method and compared using the log-rank test. The putative prognostic factors for stent patency were analyzed by the multivariate Cox proportional hazard model with forward stepwise selection. The following potential prognostic factors were included in this model: age, sex, patient performance status, underlying malignancy, stage of malignancy, obstruction site, stent type, length and diameter of stent, chemotherapy before and after stent placement, and biliary drainage before stent placement. The Factors with substantial impacts (P < 0.2) in the univariate analysis were subsequently evaluated with multivariate analysis. All statistical analyses were conducted using the IBM SPSS Statistics software, version 20 (IBM Corp., United States). A value of P < 0.05 was considered significant.

SEMS placement was conducted in 67 patients with malignant duodenal obstruction. Twenty-nine of the patients were male. Our patients were aged 68.7 ± 10.5 years (mean ± SD; range, 37 years-93 years). The causes of malignant obstruction were pancreatic cancer (46.3%), bile duct cancer (17.9%), gallbladder cancer (13.4%), ampulla of Vater cancer (7.5%), duodenal cancer (6%), others cancer including colon, bladder, and cervical cancer (6%), and hepatocellular carcinoma (3%). The median duration from tumor diagnosis to duodenal intervention was 116 d (range, 1 d-2930 d).

The baseline characteristics did not differ significantly between the 2 groups (Table 1). Of the 67 patients, 38 patients (56.7%) received uncovered SEMS, whereas 28 (41.8%) received covered SEMS, with the exception of one case of technical failure (n = 1). Three patients in the uncovered stent group required 2 overlapping stents because of a long stricture at the time of diagnosis. Before the duodenal procedure, a considerable number of patients [29 (76.3%) and 15 (51.7%) in the uncovered and covered stent groups, respectively] had undergone biliary drainage (percutaneous transhepatic biliary drainage, endoscopic retrograde biliary drainage, and surgery). Sixteen (23.9%) of the 67 patients received various palliative chemotherapeutic agents after metal stent placement. Twelve patients with advanced pancreatic cancer were administered the single agent gemcitabine (n = 2) or gemcitabine-based combination chemotherapy (n = 12). Two patients with gallbladder cancer were treated with a combination of gemcitabine and cisplatin. One patient with intrahepatic cholangiocarcinoma was managed with a combination of TS-1 and cisplatin. One patient with hilar cholangiocarcinoma was managed with a combination of gemcitabine and cisplatin.

The technical success rate was not different between the uncovered and covered stents (100% vs 97%; Table 2). Of the 29 patients who received uncovered stents, one patient experienced technical failure consisting of an inability to pass the guidewire through the obstruction of the third portion of the duodenum caused by pancreatic cancer. This patient ultimately underwent covered stent reintervention 3 d after the first intervention.

Stent placement was clinically successful in 34 (89.5%) of the 38 patients in the uncovered stent group (Table 2). The remaining 4 patients (10.5%) experienced clinical failure of the stent placement due to poor expansion (n = 2), additional obstruction (n = 1), and poor medical condition due to septic shock (n = 1). In contrast, in the covered stent group, stent placement was clinically successful in 24 (82.8%) of the 29 patients. The remaining 4 patients (13.8%) experienced clinical failure of stent placement due to stent migration (n = 2), perforation (n = 1), and septic shock (n = 1). The clinical success rate did not ultimately differ significantly between the uncovered and covered stent groups (89.5% vs 82.8%; Table 2). The GOOSS scores in the patients with technical success improved significantly compared to the scores before stent placement (mean, 1.55 vs 0.25, P < 0.001; Table 2).

During the follow-up period after stent insertion, 12 (31.6%) of the 38 patients in the uncovered stent group and 12 (41.4%) of the 29 patients in the covered stent group experienced various types of complications, including stent migration (n = 6), tumor ingrowth (n = 7), bowel perforation (n = 2), biliary obstruction (n = 3), stent collapse (n = 3), poor stent expansion (n = 2), and food impact (n = 1; Table 2). The overall complication rate did not differ significantly between the uncovered and covered stent groups (31.6% vs 41.4%; Table 2). Stent migration was more frequent in covered stents than in uncovered stents [20.7% (6/38) vs 0% (0/29); P = 0.005; Table 2], which was managed with restenting. Tumor ingrowth occurred in the uncovered (n = 6) and covered stent groups (n = 1) and was retreated with restenting (n = 5), surgical operation (n = 1), and no retreatment (n = 1). Perforation was found in 2 patients who underwent a covered stent placement, with one patient undergoing surgery and another patient refusing further treatment. Biliary obstruction developed only in the uncovered stent group without previous biliary drainage (n = 3), which was managed with percutaneous transhepatic biliary drainage (n = 3). Moreover, the retreatment rates for the uncovered and covered stents were not significantly different (31.6% vs 34.5%; Table 2).

The median follow-up was 70 d (range, 8 d-827 d). The overall cumulative median duration of stent patency was significantly longer in the uncovered stent group (251 d; 95%CI: 149.8 d-352.2 d) than in the covered stent group (139 d; 95%CI: 45.5 d-232.5 d; P = 0.033; Figure 1). The overall cumulative median survival period was not different between the uncovered stent (70 d; 95%CI: 45.8 d-94.3 d) and covered stent groups (60 d; 95%CI: 32.3 d-87.7 d; P = 0.673).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Overall stent patency between the uncovered (solid line) and covered stent groups (dotted line; P = 0. 033).

The statistical analyses of the potential factors predisposing to stent patency are summarized in Table 3. Multivariate Cox regression analysis identified that the type of inserted stent (covered) was the only independent prognostic factor for stent patency (HR = 2.586; 95%CI: 1.046-6.388; P = 0.040).

Malignant duodenal obstruction is a terminal and catastrophic complication of advanced cancers. Although the endoscopic placement of a self-expandable metal stent (SEMS) is safe and effective for palliative treatment in patients with gastric outlet obstruction, the clinical outcomes between uncovered and covered SEMS placements have not been well evaluated in malignant duodenal obstruction.

Most previous studies comparing clinical outcomes between uncovered and covered stent placement have enrolled either patients with malignant gastric outlet obstruction, mainly due to gastric cancer, or patients with malignant duodenal obstruction only as a part of their studies. The authors focused on patients with malignant duodenal obstruction.

A unique finding was that uncovered stents may be more resistant to stent migration and provide longer stent patency than covered stents.

The authors suggest that uncovered stents may be preferred when SEMS placement is performed in patients with malignant duodenal obstruction.

SEMSs are made of nitinol alloy with polytetrafluoroethylene and have a silicone covering or no covering.

This is a very interesting comparative observational study on the clinical outcomes of stent placement (uncovered vs covered) in patients with malignant duodenal obstruction. From the results of this study, practitioners can identify that uncovered stents may be preferable over covered stents.