Large bowel obstruction is a surgical emergency. It is most commonly caused by colorectal cancer and rarely by benign strictures and extrinsic compression of the colon.1 Presently, the standard of care for such patients is either a two-step diverting colostomy with subsequent colonic resection or a one-step resection with primary anastomosis with or without on-table lavage, especially in right-sided tumours where surgery is considered to be curative.
Permanent colostomy is performed in incurable, advanced cancers. Insertion of a self expanding metal stent (SEMS) is now a well established alternative to emergency surgery. It can be used both as a bridge to elective surgery, allowing time for optimization of preoperative care or as a definitive procedure in cases of locally unresectable tumours or in patients with comorbidities and therefore at high risk for surgery.
SEMS, therefore, avoids emergency surgery which has a high mortality and complication rate. Colostomy care also entails increased costs and a lower quality of life.2 However, SEMS are costly (about $ 2000/stent) and are also associated with procedure-related complications including bleeding, perforation, stent occlusion and migration.3
Over the last decade, there has been increasing international literature on the use of SEMS in acute large bowel obstruction. To date, New Zealand data has been limited to a single study which compared SEMS with traditional surgery in patients with metastatic tumours.4 We describe our own experience in a New Zealand unit with SEMS, both as a palliative procedure in inoperable cases and as a bridge to elective surgery.
Bridge to surgery
Male : Female ratio
Right-sided lesions (n)
Left-sided lesions (n)
Mean tumour length (cm)
Mean hospital stay (days)
Twenty-five patients had technically successful stent placement (90 %). Three of them required balloon dilatation of either the proximal or distal end of the stent after placement due defective expansion of the stents. One patient had a technical failure as the obstruction site was not reached and the patient was sent for emergency surgery. One patient had clinical failure after technically successful stent deployment and had emergency surgery. One patient had defective stent opening and after failure of balloon dilatation required the insertion of a second stent. One patient had immediate procedure-related perforation and died on the second day of stent placement. The three technical failure cases were from the group referred for palliative stenting.
One patient had delayed perforation 6 weeks after the placement of stent. One patient had significant haemorrhage post stenting requiring transfusion but settled on conservative treatment. Two patients had stent migration, one with diverticular stricture and another one with malignant rectal stricture one week post stent deployment. They went for elective surgery. One patient had two stents inserted to cover a long malignant stricture. Another patient had a second stent insertion after the first one migrated distally.
The mean length of hospital stay for successful stenting was 2 days (mean 1–7 days). Four cases were done as day cases and twelve were discharged the day after the procedure. Two patients died as inpatients after the procedure, one with perforation and another one with clinical failure after stent deployment. One patient had an inpatient surgery after the successful deployment of stent.
Ten patients are still alive post stenting. All except one of them had stenting as a bridge to surgery as shown in Table 1. The mean survival, to date, for them is 14 months.. Those who died after clinically successful placement of stent had a mean survival of 2.7 months (range; one week – six months).
As shown in Table 1, only one patient in the palliative group is still alive. He had resectable tumour at the time of diagnosis but surgery was precluded because of medical comorbidities . Only two patients had a survival of less than a month after successful deployment of stent.
The first reported case of colonic SEMS was by Dohmoto et al in 1991 as palliative treatment in a case of malignant stenosis of the rectum.6 Since then it has been increasingly used as a therapeutic modality not only in primary colorectal cancers but also for extrinsic and benign causes of large bowel obstruction.1,7
Colonic SEMS may be used as a primary form of palliation in advanced and unresectible tumours or as a bridge to elective surgery in patients presenting with acute large bowel obstruction. Colonic SEMS compares favourably with emergency surgery in terms of mortality and morbidity, length of hospital stay, cost-effectiveness and quality of life.
We use a combined endoscopic and fluoroscopic method in all of our patients as compared to a pure radiological method employed in some centres.8 In literature the technical and clinical success rate is around ninety percent which are comparable to our study.9
Three of our patients (10%) had defective expansion of either the proximal or distal end of the stent as picked by a follow up X-ray on the second day of the procedure. They were either successfully balloon dilated (two cases) or re-stented (one case).
Seven of our patients (25%) had either a hepatic flexure tumour (3 cases) or a transverse colon lesion (4 cases). Although, the literature suggests more technical difficulties in proximal colonic lesions all of our right-sided stents were a technical and clinical success.10 Also, once the lesion was reached we had successful guidewire cannulation of the stricture in all cases. This is better than those reported from other centres.11
We think our higher cannulation rate may be related to the use of sphincterotomes for tumours which are present on colonic bends and may be difficult to cannulate with a straight cannula or a wire. The angulation of a sphincterotome can be controlled in a graded manner and the tip directed to the axis of the tumour.
We employed uncovered stents in all of our cases except one as they less likely to migrate.12 Two of our patients had stent migration after technically correct placement and one had distal migration after misplacement (10 %). Inappropriate patient selection as those with low grade obstruction and chemotherapy given post stent insertion with shrinkage of tumour mass are other causes of stent migration in literature.13
Obstruction of SEMS with tumour in growth and over growth can happen and is commonly treated with placement of a second stent through the first stent.9,14 We have not encountered them in our study.
Bleeding is rare in SEMS placement.15 One of our patient developed significant bleeding requiring transfusion but it settled on conservative management. Perforation is more common when pre-dilatation of the stricture is performed.9 We did not routinely pre-dilate but had one case of early and one case of late perforation. Both cases did not have rescue surgery and died with comfort care. Our study, therefore, has a procedure-related mortality of seven percent.
SEMS are more cost effective when compared with emergency surgery.16 Our study with a mean hospital stay of 2 days is comparable with short hospital stay following successful placement of SEMS. This is also supported by the previous New Zealand study which showed significantly reduced mean hospital stay in the stented patients.4 We recommend SEMS deployment as a day-stay procedure in cases of uncomplicated SEMS insertion.
There is a clear role of SEMS as a definitive procedure in locally unresectable tumours presenting with acute large bowel obstruction. However, its role, as a bridge to surgery, in patients with a potentially curative disease is still debatable, with conflicting reports in literature. Clearly, the higher morbidity and mortality rates of emergency surgery, especially in left-side tumours, has to be balanced against the potential stent-related complications. Perhaps, the local expertise in each hospital and tumour characteristics (left versus right-sided tumours) should dictate the choice of intervention.
The limitation of our study was the retrospective design of the study. Data collected from the review of case notes has a potential for under-reporting of the complication rates.