Endoscopic treatment with self-expanding metal stents for Crohn's disease strictures


Correspondence to:

Dr C. Loras, Department of Gastroenterology, Endoscopy Unit, Hospital Mútua de Terrassa, Universitat de Barcelona, Plaça Dr Robert no. 5, 08221 Terrassa, Barcelona, Catalonia, Spain.

E-mails: cloras@mutuaterrassa.es;




Balloon dilation (with or without steroid injection) is the endoscopic treatment of choice for short strictures in Crohn's disease (CD). The placement of a stent has only rarely been reported in this setting, and it may be a good alternative.


To describe the efficacy of temporary placement of a self-expanding metallic stent (SEMS) in the endoscopic treatment of symptomatic strictures in CD.


We included 17 CD patients treated with SEMS (4 partially covered SEMS and 21 fully covered SEMS) for symptomatic strictures refractory to medical and/or endoscopic treatment.


We placed 25 stents in 17 patients with stenosis (<8 cm), in the colon and in the ileocolonic anastomosis. In two cases, two stents were placed in the same endoscopic procedure. All except three cases had previously been unsuccessfully treated with endoscopic dilatation. The stents were maintained for an average of 28 days (1–112). The treatment was effective in 64.7% of the patients after a mean follow-up time of 60 weeks (5–266). In four cases, removal of the stents was technically difficult due to stent impaction (moderate adverse events-AEs) and one patient had a proximal stent migration requiring delayed surgery (severe AE).


The placement of self-expanding metallic stent in Crohn's disease maintained over a period of 4 weeks is a safe, effective treatment for strictures refractory to medical treatment and/or balloon dilatation, and might be an alternative endoscopic treatment in these patients.


Stenosis is one of the most frequent local complications in Crohn's disease (CD), occurring in one of three patients within 10 years of diagnosis.[1] It happens as a result of chronic transmural inflammation with subsequent tissue remodelling, and mesenchymal cell hypertrophy due to hyperplasia and fibrosis. It occurs most frequently in the terminal ileum, ileocolonic anastomosis and rectum, and is considered clinically relevant when there is a steady narrowing of the lumen with prestenotic dilatation and obstructive symptoms.

For those who require ileal resection, more than 50% need a repeat operation after 15 years, and over 40% have recurrence of obstructive symptoms after 4 years, with the possibility in the long term of ending up with short bowel syndrome.[2]

Strictures are more common in CD than in ulcerative colitis, and they are more frequent when the disease is localised in the small bowel rather than in the colon exclusively (64% vs. 5%, respectively). In a meta-analysis which considered 516 patients with CD and strictureplasty, 36% of stenoses were found at the jejunum and ileum, 32% at the ileum and 24% at the jejunum, with recurrence of stenosis in the site of the strictureplasty in 2.8–5% of cases.[3, 4]

This suggests the need for a conservative nonsurgical treatment; currently balloon dilatation is the endoscopic treatment of choice. Several uncontrolled observational studies have shown that balloon dilatation in selected patients is a safe, effective alternative to surgery, with an overall success rate ranging from 44% to 58%.[5-7] Information regarding efficacy and safety of self-expanding metal stents (SEMS) in the treatment of strictures in CD is very limited.[8-18] Due to the high risk of recurrence of stenosis, the aim of this study was to evaluate the efficacy and safety of self-expanding metal stents (SEMS) in the treatment of strictures in CD.

Patients and Methods


We included a total of 17 CD patients (10 women, 7 men; mean age 45.7 years, range 21–62) from six Spanish referral hospitals, who were treated with SEMS refractory to medical and/or endoscopic treatment with balloon dilatation from 2006 to 2012. A stricture was considered refractory to medical and/or endoscopic treatment when there was a need for surgical treatment to relieve the obstructive symptoms. The inclusion criteria were symptomatic strictures, with a length less than 8 cm, and B2 pattern of the Montreal Classification. The only exclusion criterion for the procedure was the coexistence of a fistulising pattern (B3 of the Montreal Classification). To have more than one stricture was not an exclusion criterion when the stenoses could be treated with more than one endoscopic procedure. Patients and disease characteristics are detailed in Table S1. Symptomatic strictures were considered as occurring when patients had abdominal pain after meals with or without vomiting, and reduced or no stools and reduced passage of flatus associated with distention. The existence of stricture was demonstrated by imaging techniques (small bowel follow-through, barium enema, computed tomography scan or magnetic resonance enterography).

This is a retrospective case series and all of the patients were identified from the prospectively registered local databases of the participating hospitals. The clinical records of these patients were reviewed.

All patients provided written informed consent prior to the procedure and the study was approved by the Ethics Committee of the leading Hospital Universitari Mútua de Terrassa. It was validated by the ethics committees of the remaining hospitals.

Methods and techniques

All the stents were inserted under fluoroscopy using the through-the-scope endoscopic technique. The enteral stents were extractible and fully (FCSEMS) or partially covered (PCSEMS). Two commercially available SEMS were used [Hanarostent Colon/Rectum (CCC); M.I.Tech, Seoul, Korea and Niti-S Colonic Stent; Taewoong-Medical, Seoul, Korea]. The diameter and length of the stents are detailed in Table 1. Complementary endoscopic techniques (local steroid injection, balloon dilatation) used immediately before or during the SEMS placement were recorded.

Table 1. Stenosis and stent characteristics
PatientStenosis locationLengthPrevious treatmentType of stentTimeOutcome
  1. FCSEMS, fully covered self-expandable metal stent; IC, ileocolonic; PCSEMS, partially covered self-expandable metal stent; w, weeks.

1Sigmoid3 cmDilation (×6)PCSEMS 20×604wRecurrence 50w
PCSEMS 20×601w → impactionRecurrence 92w
FCSEMS 20×601w → spontaneous migrationSurgery
2IC anastomosis5 cmDilation + injection steroids (×5)PCSEMS 20×804w → impactionRecurrence 130w
FCSEMS 20×802w → spontaneous migrationRecurrence 22w
FCSEMS 20×802w → spontaneous migrationNo recurrence 20w
3Colon4 cmDilation (×1)FCSEMS 24×805wNo recurrence 272w
4Sigmoid2 cmDilation (×2)FCSEMS 20×8016w → impactionNot effective
5IC anastomosis2 cmDilation (×3)FCSEMS 20×602w → spontaneous migrationRecurrence 96w
FCSEMS 20×1001 day → spontaneous migrationRecurrence 13w → Surgery
6IC anastomosis4 cmDilation (×4)FCSEMS 18×808w → impactionRecurrence 43w → Surgery
7Sigmoid3 cmDilation (×2)FCSEMS 24×605 days → spontaneous migrationRecurrence 13w → Surgery
8IC anastomosis6 cmNoFCSEMS 16×60 PCSEMS 18×90Proximal migration 12wSurgery at 12w, No recurrence
9 (Figure 1)Double-colon stenosis3-4 cmNoFCSEMS both 24×60/18×60(×2)8w → spontaneous migration (18×60)Dilation at 104w → No recurrence 64w
10IC anastomosis5 cmDilation (×1)FCSEMS 18×801w → spontaneous migration → relocation 8wNo recurrence 246w
11IC anastomosis2 cmDilation (×4)FCSEMS 16×603w → spontaneous migrationRecurrence 16w → Surgery
12IC anastomosis2 cmDilation (×4)FCSEMS 16×602 days → spontaneous migrationNo recurrence 47w
13Rectal anastomosis3 cmNoFCSEMS 18×1008wNo recurrence 77w
14IC anastomosis5 cmDilation + injection steroids (×3)FCSEMS 20×80Technical failureSurgical treatment
15IC anastomosis6 cmDilation (×2)FCSEMS 20×804w → spontaneous migrationNo recurrence 14w
16IC2 cmDilation (×1)FCSEMS 20×601w → spontaneous migrationNo recurrence 16w
17Sigmoid3 cmDilation (×1)FCSEMS 20×806w → spontaneous migrationNo recurrence 11w

All the procedures were performed under deep sedation using propofol (Fresenius Kabi, Barcelona, Spain). The procedure was considered successful if (i) the patient remained free of symptoms at the end of follow-up or at least 1 year after the procedure, and (ii) no major complications related to the procedure occurred.


Characteristics of the patients

Of the 17 CD patients included, four had the disease located in the ileum (L1), one in the colon (L2) and 12 in the ileocolonic area (L3). Eight patients were treated with combined immunosuppressants and biologics, and eight cases received monotherapy. Treatment regimens are detailed in Table S1. The duration of CD before the stent procedure was a median of 15 years (range 1–30). Thirteen patients were operated on before the stent placement. In these patients, the type of surgery was ileal resection in two cases, ileocolonic resection in 10 and subtotal colectomy in one. In all cases except three, previous endoscopic dilatations with and without steroid injection had been performed unsuccessfully, with the number of dilatations ranging from 1 to 6 (and diameter of the balloons ranging from 8 to 18 mm). Co-adjuvant local steroid injection was used in three cases (Triamcinolone acetonide, Trigon Depot 40 mg/mL; Bristol-Myers Squibb S.L (New York, NY, USA), 1 mL of 40 mg/mL diluted to 5 mL with normal saline injected in 0.5–1.0-mL in each quadrant).

The most important characteristics related to the procedure (stent placement, length of the stenosis and previous endoscopic treatment) and technical procedures (type of stent and duration of placement) are detailed in Table 1.

Technical procedure, efficacy and outcome

A total of 25 stents were placed in 17 patients. Four were PCSEMS and 21 FCSEMS. The length and diameter of the stents ranged from 6 to 10 cm and from 16 to 24 mm respectively. In two cases, two stents were placed during the same endoscopic procedure, in one case for the treatment of two colonic stenoses (Figure 1) and in the other case due to a proximal migration during the stent insertion [see adverse events (AEs) below]. The stents were maintained for a mean of 28 days (range 1–112), until they were endoscopically removed or they presented spontaneous migration.

Figure 1.

Case 9: Two stents placed at sigmoid and transverse colon.

The stents were placed correctly in 16 of 17 patients (23/25 stents, 92% technical success rate in the stent placement). The follow-up period and outcome for each patient are detailed in Table 1. The treatment was successful in 11 of 17 patients (64.7%) with a mean follow-up period of 67 weeks, ranging from 11 to 272 weeks (Figures 2 and 3; Table 1). In three cases, the follow-up was shorter than 16 weeks. The recurrence rate after the therapy in patients with technical success was 43.7% (7/16) which occurred in a mean follow-up period of 58 weeks ranging from 13 to 130 weeks. In patients failing stent placement, additional therapy (endoscopic treatment or surgery) was decided by physicians in charge. In three cases, the patients were re-stented for the treatment of recurrence, with success in two patients (66.6%).

Figure 2.

(a) Transverse colon stenosis before stent insertion, (b) colon stenosis 3 months after removal of the stent.

Figure 3.

Survival curve that represents the percentage of symptom-free patients during the follow-up. The majority of recurrences occurred in the first year of the follow-up.

There were no AEs requiring urgent surgery (perforation or bleeding) during the procedure or stent maintenance. However, based on ASGE workshop recommendations[19] there were four moderate AEs related to impaction of the stent. In fact, the removal was technically difficult for 4 of 25 stents (16%), two with PCSEMS (50% of the PCSEMS) and two with FCSEMS (9.5% of the FCSEMS). In these latter cases, the FCSEMS was maintained for 8 and 16 weeks, respectively (Figure 4), with this lengthy period probably responsible for the impaction. In another patient there were severe AEs due to a proximal migration during the placement of the FCSEMS. Endoscopic removal was not possible due to colonic angulation even though another stent was correctly inserted during the same procedure. Although the patient was asymptomatic, he required a programmed surgical operation 4 months later for stent removal.

Figure 4.

Removal of impacted stent: (a) beginning of the removal, (b and c) removal of the stent remains, (d) postremoval of the stent.

Thirteen of the 25 stents (52%) presented spontaneous distal migration, due to resolution of the stenosis in the majority of the cases (11/13, 84.6%). The migration of the stents was not considered as AEs but rather an incident that could be treated in all the patients on an out-patient basis.


To our knowledge, the present series is the largest dealing with the efficacy and safety of SEMS in the treatment of strictures in CD patients. The results obtained demonstrate that this technique may offer an alternative endoscopic treatment to balloon dilatation in refractory cases or in cases in which balloon dilatation is unsuitable. In addition, the results obtained suggest that SEMS could also be considered an alternative endoscopic treatment to balloon dilatation in some cases. However, the limited number of patients (three cases in our study) treated with SEMS as a first line endoscopic option, albeit with encouraging results, does not allow this to be considered the best initial choice for endoscopic treatment.

In the literature, a total of 12 case reports and two short series have previously been published (Table S2). In the first reported cases, this technique was considered as a bridge therapy to surgery. This fact, together with the heterogeneity of the types of stents, made it difficult to draw conclusions regarding efficacy.[8-16]

The first published series was a retrospective study of five refractory ileocolonic anastomotic strictures after CD resection using uncovered stents with success in four patients, with long-term follow-up in two cases and without stent removal.[17] The most remarkable thing about this prior study was the homogeneity of the patients included, with the same type of stent used in a similar clinical setting (anastomotic recurrence after surgery), which was probably responsible for the high rate of success.

The other series included 11 patients with intestinal strictures (two naïve to surgery and nine with ileocolonic or ileocecal resection) treated with FCSEMS. The procedure was considered completely successful (stent extraction and disappearance of obstructive symptoms) in only one case (9%), but in another three cases, the patients were free of obstructive symptoms for more than 1 year after the spontaneous migration of the stent. These cases could also be considered in fact successful treatment. In two cases (18%), major morbidity requiring urgent surgery was noted and spontaneous migration of the stent was almost the rule.[18] In our study, we also found a high percentage of spontaneous distal migration of the stent (52%), which was partially due to the resolution of the stenosis in the majority of the cases (84.6%). However, if the migration is not related to any additional negative incident, the stent can be endoscopically removed, and stent migration should not be considered a true complication but rather a consequence of the resolution of the stricture.

In only one case, with proximal stent migration occurring during stent placement, was endoscopic removal not possible, due to colonic angulation. Although the patient remained free of obstructive symptoms, an elective surgical intervention was performed 4 months later for stent removal.

In our study, the overall efficacy (64.7%) was equal to or even greater than that obtained using balloon dilatation in high-risk patients refractory to standard endoscopic treatment. The efficacy of balloon dilatation has recently been reported in a review by Hassan et al.[5] of 14 published studies that included 347 patients. The overall efficiency of balloon dilation in intention-to-treat analysis [percentage of patients without surgery with a mean follow-up of 33 months (2–168)] was 58%, with an initial technical success rate of 86% and only 2% of major complications. Two recent studies [6, 7] retrospectively evaluated the efficacy of balloon dilatation in patients treated in a single centre. The immediate postprocedural technical success rate ranged from 89%[7] to 97%,[6] without the need for further dilatations and surgery in 44%[6] and 52%,[7] and with a mean follow-up of 5 years or longer. The frequency of major complications ranged from 2.5 to 5.3%.[6, 7]

Although the SEMS technique is more expensive than balloon dilation it should be noted that a mean of 2.2 dilatations is necessary in the majority of patients to achieve a favourable outcome. Therefore, a cost-efficacy comparison, considering not only the cost of the devices but also the global cost of the procedures (including the number of admissions required, complications, etc.) of the two techniques is needed to clarify this issue. In addition, development of SEMS or enteral biodegradable stents specifically designed to treat benign intestinal strictures is needed to improve these results.

The treatment of recurrence after a stent placement is not well established. However, re-stenting is a possibility that was successfully used in two patients of this series.

Regarding the four cases with difficult removal of the stent due to impaction, it is worth noting that in two cases the stents were PCSEMS (50% of the FCSEMS vs. 9.5% of the PCSEMS) and the other two were FCSEMS maintained for long periods of time (8 and 16 weeks). The overgrowth or ingrowth of mucosal hyperplastic reaction in PCSEMS may facilitate the impaction of the stent and for this reason this type of stent is not advisable for the treatment of benign strictures. In our opinion, FCSEMS for a limited period of time—no more than 4 weeks—seems to be the better choice for the treatment of CD strictures.

The major limitations of this study are its retrospective design and the heterogeneity of the SEMS used. Although the inclusion of patients from multiple centres may introduce additional heterogeneity, the study reflects the results that may realistically be obtained in clinical practice.

In conclusion, the placement of FCSEMS in CD maintained over a period of 4 weeks is a safe and effective treatment for strictures refractory to endoscopic dilatations.


Declaration of personal interests: None. Declaration of funding interests: The ‘Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd)' is an initiative of the Instituto de Salud Carlos III, Spain.