Acute large-bowel obstruction represents a surgical emergency; in 85% of cases, it is caused by the presence of a colorectal tumor (1). It has been reported that ∼15-20% of patients with colorectal cancer present with an emergent obstruction of the large bowel at the time of diagnosis (2).
Two decades ago, urgent surgery was the first choice of treatment for these patients. Such patients are usually in a poor general condition because of the underlying tumoral disease, dehydration and electrolyte imbalance, and friable colon mucosa due to large-bowel distension (3,4). Because of these factors, emergency surgery is associated with high morbidity and mortality rates (5) and very often includes a colostomy 6 which becomes permanent in 40% of patients (7,8) and leads to psychological distress (9). Although the relevance of preoperative colonic decompression has recently been questioned by some authors (10,11), elective surgery offers better results (3), with morbidity and mortality rates of 23% and 3.5%, respectively (12) vs. 39% and 12% in emergency surgery cases.
The main goals of emergency endoscopic treatment are to avoid emergency surgery and stoma creation. In patients who are surgical candidates, the placement of both decompression tubes and self-expanding metal stents (SEMSs) has proved to be effective in restoring large-bowel transit, allowing patient preparation for elective surgery and tumoral disease staging (13,14). However, rectal tubes are effective only in distal left colon tumors and are associated with significant patient discomfort and the need for repeat intervention (15). SEMSs have the advantage of being completely internalized and expanding to create a wider lumen than that achieved with tubes.
A number of stents have been designed specifically for use in the lower gastrointestinal tract and are available in a variety of lengths and diameters. Several articles have been published reporting on the efficacy and safety of different types of SEMSs as a bridge to surgery procedure in patients with potentially resectable colorectal cancer (4,16,17). The aim of the present study was to evaluate the efficacy and safety of the WallFlex colonic stent (Boston Scientific, Natick, MA) as a bridge to surgery in patients with malignant large-bowel obstruction.
In this prospective multicenter international study, patients were enrolled in two identically structured registries: the Colonic International Registry and the Colonic Spanish Registry. There were 39 participating centers from 13 different countries in 22 academic and 17 community hospital settings.
The registries included patients with colorectal strictures secondary to malignant neoplasms requiring either palliative or bridge-to-surgery stenting because of the presence of obstructive bowel symptoms (inability to pass flatus or stool in conjunction with bowel distension). However, this article reports only on the cohort with bridge-to-surgery indication. Exclusion criteria included placement of a previous colonic stent, enteral ischemia, suspected or impending perforation, intra-abdominal abscess, very poor general condition due to severe comorbidity, or a terminal stage of the patient's tumoral disease, and any use of the stent other than those specifically outlined under indication for use.
Patient data were collected from stent placement and throughout follow-up until study exit. The following baseline characteristics were documented: patient demographics, symptoms of colonic obstruction, ASA (American Society of Anesthesiologists) scoring, tumor characteristics (including origin and location), and procedural details such as the number and size of inserted stents. Patients from the bridge-to-surgery group were followed until elective surgery or until there was compromised stent functionality requiring re-treatment. All follow-up visits were conducted per standard of practice at each center. Complications or adverse events were collected and documented throughout the entire length of the study. The number of days from the time of stenting to elective surgery was also recorded.
Three analysis cohorts were used in the study. The procedural success cohort consisted of all enrolled bridge-to-surgery patients who received a stent. Data from this cohort were used to report on baseline characteristics, procedural success, and safety. The clinical success cohort consisted of patients who received a stent and had procedural success and available follow-up data through elective surgery. This cohort was used to evaluate effectiveness measures. The surgery cohort consisted of patients who had had surgery. This cohort was used for assessment of surgical outcomes.
Outcomes measures were defined as follows:
- Procedural success was reported by the treating physicians as a measure of successful endoscopic placement of the stent in the correct position.
- Clinical success—the primary effectiveness measure—was defined as having adequate passage of stool from stenting until surgery without any stent-related complications and without the need for endoscopic reintervention or emergent surgery. Passage of stool was characterized as adequate (excellent, good, or fair) or poor.
- Safety was assessed based on stent-related complications and adverse events reported in two categories: procedural complications (up to 6 h of stent placement) and post-procedural complications (from 6 h after stent placement up to surgery).
- Evaluation of surgical outcome took into account the ability to perform elective surgery as planned, with or without the need for a diverting stoma. Complications after surgery were also documented.
Device and stenting procedure
A commercially available expandable uncovered metal colonic stent that passes through the working channel of a standard adult colonoscope with midbody diameters and proximal flange diameters of 22/27 and 25/30 mm, respectively, and with lengths of 6, 9, and 12 cm (WallFlex, Boston Scientific) was used.
The stents were placed endoscopically using fluoroscopic guidance as previously described (18) (Figure 1). The procedure was performed under sedation or general anesthesia according to the usual clinical practice of each participating hospital. Prophylactic antibiotics were not administered before stent placement. Stricture dilation before stent placement was allowed at the discretion of the endoscopist.
Summary statistics were computed for the enrolled population. For categorical measures at each visit, summary statistics consisted of frequency and percentage of responses in each category. Unless otherwise noted, the denominator of a percentage is the number of subjects with nonmissing values, based on available follow-up data. For continuous measures at each visit, summary statistics include sample size, mean, median, s.d., minimum, and maximum. The sample size is the number of nonmissing values, based on evaluable patients, unless noted otherwise.
A total of 182 patients were referred for colorectal stent placement with the indication of bridge-to-surgery stenting. Of the 182 patients, 121 were enrolled in the Spanish Registry and 61 in the International Registry. All enrolled patients received a stent.
Of the 182 patients, 110 (60.4%) were male and mean (s.d.) age was 70.1 (11.2) years. The reported ASA score was I in 36 (19.8%), II in 108 (59.3%), III in 34 (18.7%), IV in 2 (1.1%), and unavailable in 2 (1.1%) patients.
All patients presented with colonic obstruction symptoms. Colorectal cancer was reported in 98.3% of patients, and the remaining 1.7% of cases had an extracolonic cancer. The tumor was located in the left colon in 84.6% of patients, in the rectum in 11% of patients, and proximal to the splenic flexure in 4.4% of cases. At initial tumoral staging, 85.7% of patients had localized colorectal cancer without metastatic disease (Table 1).
All patients were prepared with cleansing enemas to facilitate stricture visualization and further stent insertion. Stenting was performed by 35 gastroenterologists and 5 surgeons among the two registries. A colonic stent was placed with the aid of fluoroscopy and under direct visualization with the colonoscope. Sedation was employed during 129 procedures (71.3%), general anesthesia in 4 procedures (2.2%), and no sedation was used in 48 cases (26.5%). In only 5 patients (2.7%), the stricture was dilated before stent placement using either a balloon or a bougie dilator.
In 111 patients (61.3%), a stent 9 cm long was chosen for stenting, with 25/30 mm body/flare diameter in 73 patients (40.3%) and 22/27 body/flare diameter in 38 patients (21%). Stents of 6 and 12 cm length were less commonly used. Because of the length of the stricture, it was decided to place two stents to bridge the lesion in 6 cases (3.3%), whereas in the remaining patients (96.7%) placement of one stent was considered enough to treat the stricture.
The stent could be released in 181 patients; for the remaining patient who also received a stent, there were insufficient data to evaluate procedural success (Figure 2a). Procedural success was achieved in 97.8% (177/181) of stented patients. Technical failure occurred in 4 patients, because of incomplete stent deployment in 1 case and stent malposition in 3 cases. These patients were included in the assessment of safety but were excluded from the assessments of effectiveness.
Evaluation of clinical success was possible in the 150 patients who constituted the clinical success cohort. As illustrated in Figure 2a, 32 enrolled patients were not included in this evaluation. Patients were excluded for the following reasons: “crossover” to palliative care (11), lost to follow-up (12), death resulting from progression of disease before elective surgery (2), procedural failure at the initial stent placement (4), missing information regarding passage of stool (2), and lack of data beyond demographic information (1).
Clinical success, as defined in the Methods section, was achieved in 94% (141/150) of patients.
Of all enrolled patients, 157 underwent surgery (Figure 2b). Elective surgery was performed in 150 patients (89.8%), all of whom had adequate passage of stool at the time of intervention. Four patients who had elective surgery did not achieve clinical success according to our definition of the primary effectiveness measure; therefore, clinical success in the elective surgery group was achieved in 146 patients. Seven patients (4.2%) experienced emergency surgery as described in the safety section below.
Stent indwell data were available for 142 of 157 patients with surgery, and the median time from stent placement to surgery was 14 days (interquartile range, 8-20 days). During surgery, the stent was observed to be in situ except in one patient who presented with anal stent migration that had been clinically silent during follow-up. The tumor could be resected in all cases. In patients with the stent in situ, stents were removed en bloc with the tumor. A stoma was performed in 9 of the 150 patients (6%) because of surgical difficulties in performing primary anastomosis in 2 patients with rectal tumors and for reasons not specified in the remaining 7 patients.
Postsurgical complications occurred in 25 of the 150 patients who underwent elective surgery (16.6%), including both surgical and medical complications that presented during post-elective surgery hospital stay. Of these 25 patients, 3 died, 2 because of medical complications (cerebral insult and lung complications) and only 1 as a result of surgery-related complications (abdominal abscess).
Procedural complication rates are reported for all 182 enrolled patients. Post-procedural complication rates are reported for 167 patients, excluding 1 patient for whom no data were available beyond baseline demographics, 2 patients who died because of progression of disease, and 12 patients who were lost to follow-up before elective surgery.
Major procedural complications related to the stent occurred in 6 patients (3.3%), and minor adverse events (transient abdominal pain) were experienced in 2 patients (1.1%). The major complications were perforation in 3 patients (1.7%), requiring emergency surgery; persistent obstruction in 2 patients (1.1%), also surgically treated; and transient bleeding in 1 patient (0.5%) that was self-limiting.
Post-procedural complications occurred in 7 patients (4.2%) and included 2 cases (1.2%) of stent migration, resulting in endoscopic removal of the stent and restenting in one patient and in spontaneous solution through passing the migrating stent in the other; 2 cases (1.2%) of perforations that were surgically treated; 1 patient (0.6%) with persistent obstruction who required a second stent; and 2 cases (1.2%) of stent occlusion due to fecal impaction that were endoscopically managed without any further reintervention. One of the patients (0.6%) who presented with a colonic perforation 6 days after stent placement underwent emergency surgery and experienced further complications after surgery (sepsis), which was followed by death 6 days later. Concerning minor post-procedural adverse events, only 1 patient (0.6%) presented with pain, which was successfully treated with standard analgesia.
Considering both procedural and post-procedural major complications, the overall complication rate was 7.8% (13/167). A summary of the types of complications, times of occurrence, and management of the complications in accordance with the details provided above can be found in Table 2.
During follow-up, six patients (3.3%) died, as previously described: three after surgery, one because of complications related to surgery administered for the management of a perforation, and two owing to progression of the disease before surgery.
Colorectal SEMSs have revolutionized the treatment of patients with acute malignant colonic obstruction (19). Stenting can convert an emergent one- or two-step surgical procedure including a colostomy into an elective one-step resection with successful primary anastomosis rates at least twice as high as those after emergency surgery and with a remarkable decrease in number of colostomies (20,21). Furthermore, placement of the SEMS allows the treating physician to perform a complete preoperative colonoscopy to exclude synchronous lesions (22) and to perform accurate tumoral staging—essential for determining the best therapeutic option for the patient, considering that up to 50% of patients who undergo emergency laparotomy are not candidates for curative surgery (3).
In this largest series to date of bridge-to-surgery stenting, procedural success was achieved in 97.8% of patients, results consistent with previously published data, notably, 83% in a recent prospective study of 30 patients by Brehant et al. (23), 91.9% (range 33-100%) from a pooled analysis of 21 studies (n=407) by Sebastian et al. (13), and 94% in a study by Vemulapalli et al. (24). A much lower technical success rate of 70% (33/47) was reported by van Hooft et al. (25). An editorial article (26) speculates what the reasons for this low technical success rate might be.
Our clinical success rate was 94%, which is in line with that in other publications on colonic stenting as a bridge to surgery, namely, 71.7% (range 45-84%) from the pooled analysis by Sebastian et al. (13), 85% by Khot et al. (27), and 94.7% by Repici et al. (17). Of note, the definition of clinical success in our study—achievement of adequate passage of stools from stenting to elective surgery without endoscopic or surgical reintervention—truly describes the main purpose of bridge-to-surgery stenting. The 11 “crossover” patients who during their follow-up became unfit for surgery were excluded from the clinical success analysis despite the fact that improvement in passage of stool alone was probably a clinical benefit from stenting. Given that a significant portion of patients operated on an emergency basis cannot be cured because of advanced tumoral disease, the avoidance of urgent surgery is a valuable contribution of stenting in such deteriorated patients with short life expectancy.
Our overall complication rate was 7.8% (13/167), with 3% (5/167) perforations and 0.6% (1/167) stent-related mortality, which is in line with published overall complication rates (5.3-5.9%) (13,17,28), perforation rates (3.76%) (13), and mortality rates (0.4%) (13) and significantly lower than published emergency surgery outcomes (28,29,30) Two studies by van Hooft et al. (31,25) reported much higher rates of perforation (13% and 66.6%, respectively) than described elsewhere in the literature and seen in our own experience. The authors suggested that factors contributing to the high perforation incidence might have been stent design and chemotherapy (31) and the selection of a patient population with an increased risk of complications, or that perhaps the short interval (5-16 days) between stent placement and elective surgery did not allow for optimal bowel decompression and improvement in the clinical condition of the patient (25). Our elective surgery morbidity and mortality rates are consistent with previously published data with results ranging between 0 and 22% and between 0 and 5%, respectively (5,8,16,30,32). Our morbidity results included both surgical complications and all medical adverse events during the postsurgery hospital stay, thus making our rate higher than that in reports by others who only considered anastomotic dehiscence (32). A similar comment can be made about our mortality rate, as two of the three patients who died after surgery had fatal medical complications (cerebral insult and aspiration pneumonia) and only one died because of a surgery-related complication (abdominal abscess and further sepsis).
Bridge-to-surgery stenting allowed elective surgery with successful primary anastomosis to be performed in 150 (89.8%) of all evaluable patients in the colonic registries. The stoma performance rate was 6.6% (12/182), with only 9 patients (6%) requiring a diverting stoma and 3 patients an emergency stoma because of a complication. Such stoma-avoidance advantage of stenting in comparison to emergency surgery (33,34) has been widely demonstrated with the exception of a study by Pirlet et al. (35) in which there was no statistically significant difference in stoma rate between patients who received a SEMS (43%) and those who did not (57%) prior to surgery. However, all stomas from the SEMS arm were done in technical-failure patients who actually underwent emergency surgery, whereas the majority of successfully stented patients had the benefit of elective colonic resection and primary anastomosis with an uneventful postoperative course. Therefore, our stoma creation results can be emphasized because they are considerably lower than other published data (30,34,35).
Special attention should be focused on rectal tumors. Sixteen patients with rectal cancer underwent elective surgery in our study, only 2 of whom (12.5%) required a stoma to protect a primary anastomosis. This is approximately twice the overall stoma-formation rate of 6.6% in our study. Other authors (30,34) report even higher stoma rates (45% and 85.7%, respectively) in rectal tumor patients. In some studies, rectal cancer patients were excluded in consideration of differences in preoperative chemoradiotherapy and the significant technical difficulty of rectal resective surgery (18). Further studies might be needed to evaluate the effectiveness and safety of bridge-to-surgery stenting in patients with rectal tumors.
The authors recognize three principal limitations of the study. First, the nonrandomized study design does not allow for comparison with other treatment options. Second, clinical success, corresponding to the primary outcome, was reported for 150/182 patients. Of the 32 cases for which primary outcome could not be assessed, 17 were related to reported procedural failure (4 patients) and changes in the patient's medical condition (13 patients), such as death due to progression of disease before surgery in 2 patients and transition to palliative treatment in 11 patients. The absence of primary outcome data for the remaining 15 patients (8.2%), however, was the result of shortcomings in clinical trial conduct: 12 patients (6.6%) were lost to follow-up, data regarding passage of stool were missing for 2 patients, and all primary end point data were missing for 1 patient (see Figure 2a). Third, the date of surgery was missing for 15 (8.2%) patients in the surgery cohort resulting in the reporting of median stent indwell surgery for only 142 patients.
Considering our results, we can conclude that colonic SEMSs are effective in patients with acute malignant colonic obstruction as a bridge-to-surgery treatment, restoring luminal patency and allowing elective surgical resection with primary anastomosis. Moreover, the use of this stent is safe and associated with an acceptable complication rate.
The Colonic Registries were sponsored by Boston Scientific Corporation (BSC). We gratefully acknowledge the help of BSC employees Terry Liao, John Evans, Brian Johnson, and Mathew Rousseau for statistical analyses; Lina Ginnetti, Joy Peetermans, and Medical Monitor Robert Walsh for editorial assistance with this article; and Montserrat Agustí and Eduardo Sessa for planning and execution of the Registries.
CONFLICT OF INTEREST
Guarantor of the article: J. Jiménez-Pérez, MD.
Specific author contributions: J. Jiménez-Pérez: planning and conducting the study, collecting and interpreting data, and drafting the manuscript; J. Casellas, J. García-Cano, J. Vandervoort, O. Roncero García-Escribano, J. Barcenilla, A. Alvarez, P. Goldberg, F. Gonzalez-Huix, and E. Vázquez-Astray: conducting the study and collecting data, and approving the manuscript; S. Meisner: planning and conducting the study, collecting and interpreting data, and approving the manuscript.
Financial support: Boston Scientific sponsored the study in the form of financial support, study design, statistical analysis, and editorial assistance.
Potential competing interests: S. Meisner: consultant for Boston Scientific/Coloplast Denmark; J. Jiménez-Pérez, J. Casellas, J. García-Cano, J. Vandervoort, O. Roncero García-Escribano, J. Barcenilla, A. Alvarez, P. Goldberg, F. Gonzalez-Huix, and E. Vázquez-Astray: None.
1. Pommergaard HC, Vilmann P, Jakobsen HL et al
. A clinical evaluation of endoscopically placed self expanding metallic stents in patients with acute large bowel obstruction. Scand J Surg 2009;98:143-147.
2. Fan YB, Cheng YS, Chen NW et al
. Clinical application of self expanding metallic stent in the management of acute left sided colorectal malignant obstruction. World J Gastroenterol 2006;12:755-759.
3. Soto S, López-Rosés L, González-Rámirez A et al
. Endoscopic treatment of acute colorectal obstruction with self expanding metallic stents. Surg Endosc 2006;20:1072-1076.
4. Small AJ, Baron TD. Comparison of Wallstent and Ultraflex stents for palliation of malignant left sided colon obstruction: a retrospective, case-matched analysis. Gastrointest Endosc 2008;67:478-488.
5. Dulucq JL, Wintringer P, Reyssac R et al
. One stage laparoscopic colorectal resection after placement of self-expanding metallic stents for colorectal obstruction. A prospective study. Dig Dis Sci 2006;51:2365-2371.
6. Scott NA, Jeacock J, Kingston RD. Risk factors in patients presenting as an emergency with colorectal cancer. Br J Surg 1995;82:321-323.
7. Watt AM, Faragher IG, Griffin TT et al
. Self-expanding metallic stents for relieving malignant colorectal obstruction. A systematic review. Ann Surg 2007;246:24-30.
8. Kim JS, Hur H, Min BS et al
. Oncologic outcomes of self-expanding metallic stent insertion as a bridge to surgery in the management of left sided colon cancer obstruction: comparison with non obstructive elective surgery. World J Surg 2009;33:1281-1286.
9. Karadag A, Mentes BB, Uner A et al
. Impact of stomatherapy on quality of life in patients with permanent colostomies or ileostomies. Int J Colorectal Dis 2003;18:234-238.
10. Van't Sant HP, Weidema WF, Hop WC et al
. The influence of mechanical bowel preparation in elective lower colorectal surgery. Ann Surg 2010;251:59-63.
11. Ciga MA, Oteiza F, Fernandez L et al
. Comparative study of one-stage colectomy of the descending colon in emergency and elective surgery without mechanical preparation. Dis Colon Rectum 2010;53:1524-1529.
12. Leitman IM, Sullivan JD, Brams D et al
. Multivariate analysis of morbidity and mortality from the initial surgical management of obstructing carcinoma of the colon. Surg Gynecol Obstet 1992;174:513-518.
13. Sebastian S, Johnston S, Geoghegan T et al
. Pooled analysis of the efficacy and safety of self-expanding metal stenting in malignant colorectal obstruction. Am J Gastroenterol 2004;99:2051-2057.
14. Adler DG, Baron TH. Endoscopic palliation of colorectal cancer. Hematol Oncol Clin North Am 2002;16:1015-1029.
15. Adler DG. Management of malignant colonic obstruction. Curr Treat Options Gastroenterol 2005;8:231-237.
16. Baik SH, Kim NK, Cho HW et al
. Clinical outcomes of metallic stent insertion for obstructive colorectal cancer. Hepatogastroenterology 2006;53:183-187.
17. Repici A, De Caro G, Luigiano C et al
. WallFlex colonic stent placement for management of malignant colonic obstruction: a prospective study at two centers. Gastrointest Endosc 2008;67:77-84.
18. Jost RS, Jost R, Schoch E et al
. Colorectal stenting: an effective therapy for preoperative and palliative treatment. Cardiovasc Intervent Radiol 2007;3:433-440.
19. Stimac D. Colonic stents for the palliation of malignant colonic obstruction. Dig Dis 2008;26:336-341.
20. Martínez-Santos C, Lobato RF, Fradejas JM et al
. Self-expandable stent before elective surgery vs. emergency surgery for the treatment of malignant colorectal obstruction: comparison of primary anastomosis and mortality rates. Dis Colon Rectum 2002;45:401-406.
21. Repici A, Conio M, Caronna S et al
. Early and late outcome of patients with obstructing colorectal cancer treated by stenting and elective surgery: a comparison with emergency surgery and patients operated without obstructive symptoms. Gastrointest Endosc 2004;59:AB275.
22. Vitale MA, Villotti G, D'Alba L et al
. Preoperative colonoscopy after self-expandable metallic stent placement in patients with acute neoplastic colon obstruction. Gastrointest Endosc 2006;63:814-819.
23. Brehant O, Fuks D, Bartoli E et al
. Elective (planned) colectomy in patients with colorectal obstruction after placement of a self-expanding metallic stent as a bridge to surgery: the results of a prospective study. Colorectal Dis 2009;11:178-183.
24. Vemulapalli R, Lara LF, Sreenarasimhaiah J et al
. A comparison of palliative stenting or emergent surgery for obstructing incurable colon cancer. Dig Dis Sci 2010;55:1732-1737.
25. van Hooft JE, Bemelman WA, Oldenburg B et al
. Colonic stenting versus emergency surgery for acute left-sided malignant colonic obstruction: a multicentre randomised trial. Lancet Oncol 2011;12:344-352.
26. Song LM, Baron TH. Stenting for acute malignant colonic obstruction: a bridge to nowhere? Lancet Oncol 2011;12:314-315.
27. Khot UP, Lang AW, Murali K et al
. Systematic review of the efficacy and safety of colorectal stents. Br J Surg 2002;89:1096-1102.
28. Park IJ, Choi GS, Kang BM et al
. Comparison of one-stage management of obstructing left-sided colon and rectal cancer: stent-laparoscopic approach vs. intraoperative colonic lavage. J Gastrointest Surg 2009;13:960-965.
29. Lee YM, Law WL, Chu KW et al
. Emergency surgery for obstructive colorectal cancers: a comparison between right-sided and left-sided lesions. J Am Coll Surg 2001;192:719-725.
30. Ng KC, Law WL, Lee YM et al
. Self-expanding metallic stent as a bridge to surgery versus emergency resection for obstructing left sided colorectal cancer: a case-matched study. J Gastrointest Surg 2006;10:798-803.
31. van Hooft JE, Fockens P, Marinelli AW et al
. Early closure of a multicenter randomized clinical trial of endoscopic stenting versus surgery for stage IV left- sided colorectal cancer. Endoscopy 2008;40:184-191.
32. Alcántara M, Serra X, Bombardó J et al
. Colorectal stenting as an effective therapy for preoperative and palliative treatment of large bowel obstruction: 9 years experience. Tech Coloproctol 2007;11:316-322.
33. Tilney HS, Lovegrove RE, Purkayastha S et al
. Comparison of colonic stenting and open surgery for malignant large bowel obstruction. Surg Endosc 2007;21:225-233.
34. Dastur JK, Forshaw MJ, Modarai B et al
. Comparison of short and long term outcomes following either insertion of self-expanding metallic stents or emergency surgery in malignant large bowel obstruction. Tech Coloproctol 2008;12:51-55.
35. Pirlet IA, Slim K, Kwiatkowski F et al
. Emergency preoperative stenting versus surgery for acute left-sided malignant colonic obstruction: a multicenter randomized controlled trial. Surg Endosc 2011;25:1814-1821.