ARTICLE IN PRESS

ORIGINAL ARTICLE

From iatrogenic digestive perforation to complete anastomotic

disunion: endoscopic stenting as a new concept of ‘‘stent-guided
regeneration and re-epithelialization’’
Laila Amrani, MD, Charles Ménard, MD, Stéphane Berdah, MD, Olivier Emungania, MD,
Phillipe Ah Soune, MD, Clément Subtil, MD, Christian Brunet, MD, Jean-Charles Grimaud, MD,
Marc Barthet, MD, PhD
Marseille, France

Background: Iatrogenic intestinal tract perforation and anastomotic disunion traditionally required surgical
treatment. Complete anastomotic break was considered until now as an absolute contraindication for endo-
scopic management.
Objective: The aim of this series was to show that endoscopic management is able to treat a spectrum of bowel
wall breaks, from focal perforation to complete anastomotic disunion.
Setting: A single-center prospective cohort study.
Patients: Nine consecutive patients with nonmalignant gastrointestinal perforations were treated with endo-
scopic stenting between 2005 and 2008. Perforations were related to endoscopic perforations (4 cases: 2 esoph-
ageal and 2 colorectal), postoperative fistula or leakage (2 cases: 1 colorectal anastomosis and 1 esophageal),
and complete anastomotic disunion (3 cases: 2 ileoanal anastomosis and 1 esophagogastric anastomosis).
Interventions: All 9 patients underwent endoscopic installation of fully covered stents under endoscopic and
radiologic guidance, sometimes associated with simultaneous endoscopic collection drainage. Oral feeding was
resumed when radiologic contrast studies showed no residual leak.
Results: The outcome in all 9 patients was favorable. Two migrated stents were replaced, and 2 stents were
spontaneously expelled without consequence. All stents were withdrawn within an average of 5 weeks.
Limitations: Uncontrolled pilot study, small sample size.
Conclusion: The successful endoscopic management of bowel wall breaks ranging from perforation to com-
plete postoperative disunion with fully covered stent could support a new concept of ‘‘stent-guided regenera-
tion and reepithelialization.’’ Controlled trials are needed before this new endoscopic treatment can be
proposed as a substitute for traditional treatments. (Gastrointest Endosc 2009;-:---.)

Digestive perforation is a dramatic and life-threatening
complication. The 2 main causes are anastomotic leakage
and endoscopy-related perforation. The criterion standard
for iatrogenic bowel perforation has traditionally been the
surgical approach. These past few years, new conservative
treatments have emerged: endoscopic clipping, biologic
glue, endoscopic sutures, and endoluminal stenting.
Abbreviations: SEMS, self-expanding metal stent; TTS, through the scope.
DISCLOSURE: All authors disclosed no financial relationships relevant
to this publication.
Copyright ª 2009 by the American Society for Gastrointestinal Endoscopy
0016-5107/$36.00
doi:10.1016/j.gie.2008.09.043
Despite all those modalities, very scarce data exists about en-
doscopic treatment of nonesophageal GI perforations.1-5
Complete anastomotic disunion was considered until now
as an absolute contraindication for endoscopic manage-
ment. The aim of this study was to show that endoscopic
management is able to treat not only partial leakage or
perforations but also complete anastomotic disunion. We
report here a case series of 9 cases of digestive perforations
caused either by an endoscopic procedure or a postopera-
tive complication between 2005 and 2007, all treated with
enteral fully covered stents.
From this series of cases of endoscopic management of
digestive perforations to complete postoperative disunion
with a fully covered stent emerges a new concept of
‘‘stent-guided regeneration and re-epithelialization.’’

www.giejournal.org Volume -, No. - : 2009 GASTROINTESTINAL ENDOSCOPY 1

 ARTICLE IN PRESS
Endoscopic stenting as a concept of ‘‘stent-guided regeneration and re-epithelialization’’
CASE REPORTS
Cases with endoscopic perforation
Case 1. A 42-year-old woman with achalasia and mega-
esophagus had esopahgeal perforation after pneumatic
dilation. A through-the-scope (TTS) 100-mm Niti-S self-
expanding metal stent (SEMS) (Taewong, Gyeonggi-do,
Korea) colonic stent with a 28-mm flared end bridged
the esophageal leak with success. Oral intake was re-
sumed on day 11, and the stent was endoscopically re-
moved after 5 weeks.
Case 2. An 83-year-old patient had a 3-cm esophageal
perforation after endoscopic resection of a superficial
squamous cell carcinoma of the esophagus with the
EMR Kit (Olympus Medical Systems, Tokyo, Japan). A
100-mm TTS Niti-S SEMS colon stent with a 28-mm flared
end was immediately put in place to bridge the esopha-
geal perforation. The patient was treated with broad-spec-
trum antibiotics for a week and discharged from the
hospital 10 days after the perforation. Five weeks later
the endoscopic control showed complete healing and
no recurrence of the squamous cell carcinoma.
Case 3. A rectal perforation occurred during colono-
scopy in a 73-year-old woman with a history of rectal ade-
nocarcinoma. A 100-mm fully covered TTS Niti-S stent was
immediately placed over the breach. The patient was ad-
mitted for observation, and broad-spectrum antibiotics
were started. Evolution was uneventful, and she was dis-
charged 48 hours later. Five weeks later the stent was
spontaneously expelled and the control endoscopy and
contrast rectogram showed no leak.
Case 4. A 52-year-old woman underwent rectal endo-
scopic ultrasonography for a suspicion of rectal wall endo-
metriosis that was complicated with a rectosigmoid
perforation. A 60-mm fully covered TTS Niti-S SEMS was
installed and antibiotics begun. Oral nutrition was
restarted, and the patient left the hospital on the twelfth
day. Three weeks later the scan showed a small residual
pheumoperitoneum and spontaneous migration of the
stent. Colonoscopy confirmed the complete healing of
the mucosa, and soluble contrast showed no leak.
Cases with postoperative fistulas or leakage
Case 5. A 34-year-old woman had postoperative peri-
tonitis, an abdominal abscess, and a cardial fistula after
a complicated Heller myotomy and fundoplicature for
achalasia. A fully covered Hanarostent SEMS (Life Partners
Europe, Bagnolet, France) was placed under radiologic
guidance. Oral intake was reintroduced 2 weeks later,
and the patient was discharged after a month’s stay. The
stent was removed endoscopically after 5 weeks with no
subsequent compli-
cation.
Case 6. A 78-year-old man with rectal adenocarcinoma
was treated with coloanal anastomosis and defunctioning
2 GASTROINTESTINAL ENDOSCOPY Volume -, No.
Amrani et al
Capsule Summary
What is already known on this topic
d Iatrogenic intestinal tract perforation and anastomotic
disunion have been considered as absolute
contraindications for endoscopic management.
What this study adds to our knowledge
d Nine consecutive patients with nonmalignant GI
perforations were treated successfully with covered
stents placed endoscopically.
d Two stents migrated and were replaced, whereas 2 stents
were spontaneously expelled without consequences.
ileostomy. After closure of the ileostomy, an infected co-
loanal anastomotic fistula was diagnosed. Over the next
weeks, multiple trials of bowel reanastomosis complicated
with relapsing pelvic abcesses ended up with bowel diver-
sion. The patient refused definitive ileostomy and abdom-
inoperianal amputation, so an endoscopic treatment was
proposed. A 100-mm fully covered TTS Niti-S SEMS was
placed with its distal end lying directly on the upper
anal margin. The stent was removed 5 weeks later, after
resolution of the fistulous tracts was confirmed.
Cases with complete anastomotic disunion
Case 7. A 42-year-old woman with ulcerative colitis un-
derwent total proctocolectomy with ileoanal anastomosis
and defunctioning ileostomy complicated by anastomotic
leakage and pelvic and retroperitoneal abscess. The lack
of improvement with surgical drainage justified an at-
tempt at endoscopic management. Endoscopy revealed
multiple fistulous leaks resulting from complete anasto-
motic disunion. A pigtail drain was endoscopically placed
into the main fistulous tract, and a 60-mm fully covered
Niti-S SEMS bridged the anastomosis (Figs. 1 to 5). Five
weeks later, the stent was removed and a secondary anas-
tomotic stenosis was dilated. The previous anastomotic
disunion site was fully covered by granulation tissue and
regenerating mucosa. Five months later the ileostomy
was removed and bowel continuity restored.
Case 8. A 23-year-old diabetic woman underwent total
proctocolectomy with ileoanal anastomosis and defunc-
tioning ileostomy for long-standing ulcerative colitis. Post-
operative follow-up was complicated by a complete
anastomotic disunion with multiple posterior fistulas com-
municating through a 4-cm collection with the anastomotic
small bowel. After heavy irrigation of the collection, a fully
covered TTS Niti-S SEMS was put in place, bridging the
lower pouch to the anus. The stent was removed
2 months later, and contrast showed no residual leak.
Case 9. A 74-year-old man with a T3N0 stenosing adeno-
carcinoma of the lower esophagus underwent chemoradia-
tion therapy before proximal esogastrectomy. Postoperative
- : 2009 www.giejournal.org
 ARTICLE IN PRESS
Amrani et al Endoscopic stenting as a concept of ‘‘stent-guided regeneration and re-epithelialization’’
Figure 1. Complete disunion of the ileoanal anastomosis (case 7).
Figure 2. Insertion of a drainage catheter in a fistulous tract and a guide-
wire in the distal ileum before the deployment of the SEMS (case 7).
Figure 3. SEMS inserted through the anal verge (case 7).
period was complicated by a complete anastomotic dis-
union. Before surgical diversion, an endoscopic approach
was attempted in agreement with the patient and a multidis-
ciplinary committee. The gastric end was lying a centimeter
away from the lower esophagus within a 3-cm-wide medias-
tinal cavity. After heavy rinsing, gastric canulation was ob-
tained by using a stiff guidewire (Jagwire, Microvasive
www.giejournal.org
Figure 4. The 5-week endoscopic control showing complete repair of
the anastomotic disunion (case 7).
Figure 5. Radiologic control during endoscopic procedure showing no
residual leakage (case 7).
Endoscopy, Boston Scientific, Natick, Mass). Under fluoro-
scopic guidance, a 100-mm fully covered colon TTS Niti-S
SEMS was installed. The stent proximal end was fixed to
the esophageal wall with 2 clips (Resolution Clip, Microvasive
Endoscopy, Boston Scientific). Injection of soluble contrast
showed no residual leak. Ten days afterward, because of peri-
prosthetic reflux of gastric content in the mediatinum and
the impacted distal end of the stent in the gastric wall,
another 100-mm fully covered colon TTS Niti-S SEMS was
placed inside the first so that its distal end lay in the antrum.
One month later, clips were removed and both stents
withdrawn. Complete re-epithelialization had occurred
(Figs. 6 to 9). Contrast showed a small fistulous tract in the
distal stomach that was obturated with a cyanoacrylate-lipio-
dol preparation (Glubran 2 [Biocorp, France]; Lipiodol Ultra
Volume -, No. - : 2009 GASTROINTESTINAL ENDOSCOPY 3
 ARTICLE IN PRESS
Endoscopic stenting as a concept of ‘‘stent-guided regeneration and re-epithelialization’’
Figure 6. Complete esophagogastric anastomotic disunion (case 9):
a guidewire is inserted into the proximal stomach.
Figure 7. Complete esophagogastric anastomotic disunion (case 9):
a fully covered stent is inserted and clipped to the mucosa, above the
anastomotic leakage.
Fluide [Guerbe, Roissy, France]), and the orifice was closed
with 2 clips. Oral intake was resumed the same day. A month
later the patient resumed chemotherapy.
RESULTS
The 9 consecutive patients, referred from digestive sur-
geons or local gastroenterologists, treated for bowel tears
were all successfully managed with fully covered enteral
stents. All the patients healed within 12 days to 2 months.
Complications encountered were 4 stent migrations in
2 esophageal perforations, 2 spontaneous expulsions of
colorectal stents in 2 patients, and 1 anastomotic stenosis
in a completely regenerated anastomotic disunion.
DISCUSSION
The consequences of iatrogenic esophageal perforation
are life threatening. In this setting, surgery is the criterion
standard but is associated with high mortality rates (12%
4 GASTROINTESTINAL ENDOSCOPY Volume -, No. - : 2009
Amrani et al
Figure 8. Complete esophagogastric anastomotic disunion (case 9):
radiologic control showing the covered stent bypassing the leak.
Figure 9. Complete esophagogastric anastomotic disunion (case 9):
endoscopic control after stent removal.
to 50%), morbidity, and prolonged hospital stays.4 In late
diagnosis there is also a 40% to 50% risk of anastomotic in-
sufficiency.2 Considering all those issues, another therapeu-
tic option is desirable. Until recently, endoscopic stenting
has only been considered when no further surgical treat-
ment is indicated.1,5 By bridging the breach, the enteral
stent serves as a guide for digestive wall healing while pre-
venting extraluminal contamination by the endogenous
bacterial flora. The sooner the breach is stented, the lesser
are the chances of significant mediastinal or peritoneal con-
tamination. Endoluminal stenting also allows resuming oral
intake sooner and permits a shorter hospital stay.1,2,4,5
Stents have shown an encouraging 82% to 94% success
rate in cases of esophageal perforations.1,4,6
The success of endoscopic treatment seems to depend
on the delay between perforation and intervention, which
www.giejournal.org
 ARTICLE IN PRESS
Amrani et al Endoscopic stenting as a concept of ‘‘stent-guided regeneration and re-epithelialization’’

TABLE 1. Characteristics of the 9 cases of perforation

Patients

No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No. 9

Endoscopic perforations Postoperative leakage Complete anastomotic

(esophageal, colorectal) or fistula disunion

Age (y) 42 83 73 52 34 78 42 23 74
Cause of Pneumatic Endoscopic Rectal Rectal Esophageal Coloanal Ileoanal Ileoanal Esophagogastric
perforation dilation for mucosal endoscopic endoscopic breach after anastomotic anastomotic anastomotic anastomotic
achalasia resection US US Heller fistula and leakage leakage leakage
(megaes (esophageal myotomy abcess
ophagus) SCC)

Stent type Hanarostent Niti-S, Niti-S, Niti-S, Hanarostent Niti-S, Niti-S, Niti-S, Niti-S,
(Life Europe) (Taewong) (Taewong) (Taewong) (Life Europe) (Taewong) (Taewong) (Taewong) (Taewong)
then Niti-S,
(Taewong)
Delay 72 h 72 h !30 min !30 min 3 wk 4 mo 4d 2 mo 20 d
between
perforation
diagnosis
and stenting
Oral intake 11 d 11 d 48 h 4d 15 d Ileostomy Ileostomy Ileostomy 3 wk
Length 12 d 12 d 48 h 12 d 5 wk 2 mo 2 mo
of stay
Complications 2 2 Spontaneous Spontaneous 0 0 Anastomotic 0 0
Migrations Migrations expulsion expulsion stenosis
Withdrawal 5 wk 5 wk !5 wk !5 wk 5 wk 5 wk 5 wk 2 mo 5 wk
of the stent
SCC, Squamous cell carcinoma.

appears to be the most important prognostic factor and
should probably be shorter than 12 hours.2,7 The second
variable is the size of the perforation. Biologic glue or en-
doscopic clipping is suggested in small perforations,
whereas endoscopic stents are used in tears going up to
70% of the circumference.8-11 In larger tears, surgery was
considered until now the only advisable approach.9,10
Covered SEMSs and self-expanding plastic stents have
been used to treat iatrogenic esophageal perforation or
postoperative leak.1,2,4,6,12 An SEMS is effective in occlud-
ing the breach, but its withdrawal is sometimes difficult
and the uncovered flared ends can cause traumatic muco-
sal lesions. Recently, new fully covered stents have been
developed, some made of metal composites, others
made out of plastic. They show the advantages of being
more easily removed or repositioned, and their full-length
coating prevents embedding into the mucosa.13,14 These
advantages make them more adapted for the treatment
of benign diseases such as digestive perforation or benign
stenosis. In this series, we mainly used Niti-S stents be-
cause of their flexible nitinol frame, fully covered transpar-
ent silicone coating, visible radiologic markers, and TTS
capability so that they could be placed with both endolu-
minal and radiologic guidance. The selection of these
stents reflects local expertise, and other stents with the
same characteristics could possibly be effective.
Fully covered esophageal stents are likely to migrate in
18% to 30% of cases.4,5,11,13,15 A radiologic control is there-
fore mandatory before oral intake is restarted and when
the condition of the patient changes. The migration risk
is lower if the stent has a larger diameter and if it is longer;
some authors have also suggested attaching the stent to
the digestive wall with clips.1,9,11 The proximally clipped
SEMS in cases 7 and 9 did not have any migration. In 2 pa-
tients the stents were placed close to the anal margin, and
they were expelled spontaneously without compromising
the overall success of the procedure.
The prosthesis should be left in place long enough so that
full re-epithelialization is allowed, but they should be re-
moved before they embed within the mucosa.15 In available
literature, the stents were left in place from 10 days up to 19
weeks.2,12,16 In our series, the stents were withdrawn after
a mean length of 5 weeks. This delay seemed to be sufficient
to heal the defect and allow removal of the stent without
damaging the mucosa.
No reported cases of complete anastomotic disunion or
nonesophageal sites treated with endoscopic stents are
described in the literature. Table 1 presents the

www.giejournal.org Volume -, No. - : 2009 GASTROINTESTINAL ENDOSCOPY 5

 ARTICLE IN PRESS
Endoscopic stenting as a concept of ‘‘stent-guided regeneration and re-epithelialization’’
characteristics and outcomes of the 9 patients discussed.
Although impressive, our 100% success rate could be
the result of an underpowered series.
Three of our cases were treated for complete anasto-
motic disunion; 2 of them were ileoanal anastomoses
and 1 was an esophagogastric anastomosis (Figs. 1 to 9).
Successful anastomotic healing was achieved in all 3 cases.
In those patients, endoscopic treatment was the last ther-
apeutic option before difficult salvage surgery. The proce-
dure also requires simultaneous endoscopic, radiologic, or
surgical drainage of infectious sites around the leaking
site, as in cases 7 and 9, which may impede proper heal-
ing. The success of endoscopic stenting in those 9 patients
could be the result of an underpowered series but shows
a potential encouraging alternative to surgical reinterven-
tion that needs to be studied in comparative trials. By this
experience, we postulate that a covered stent might act as
a support to guide tissue regeneration and re-epithelializa-
tion alongside the external membrane of the covered
stents. This was achieved in a period not exceeding
2 months in all cases. Therefore, the management of large
or complete postoperative leakage with fully covered
stents could support a new concept of ‘‘stent-guided
regeneration and re-epithelialization.’’ It is also possible
that stents actually stimulate regrowth of the mucosa,
but this remains to be proven by experimental models.
CONCLUSION
Partial to complete iatrogenic disunion can be success-
fully treated with endoscopic stents. They allow rapid res-
tauration of bowel continuity while being minimally
invasive and preserving GI anatomy and physiology.
They could also be responsible for stimulating mucosal
regrowth. Although we have presented a small sample
size, this work proposes a new concept of ‘‘stent-guided
mucosal healing,’’ which may cause rethinking of the
respective roles of surgery and endoscopy in iatrogenic
bowel perforation. A large-scale integration of this tech-
nique in current practice needs more controlled studies
and experimental series. In addition, the design of the
available stents has to be improved to better accommo-
date the various anatomic locations such as the stomach
and the colon.
ACKNOWLEDGMENT
We thank the Endoscopy Unit staff of the Hopital Nord
de Marseille for their continuous and outstanding work.
6 GASTROINTESTINAL ENDOSCOPY Volume -, No.
Amrani et al
REFERENCES
1. Freeman RK, Ascioti AJ, Wozniak TC. Postoperative esophageal leak
management with the polyflex esophageal stent. J Thorac Cardiovasc
Surg 2007;133:333-8.
2. Fischer A, Thomusch O, Benz S, et al. Nonoperative treatment of 15
benign esophageal perforations with self-expandable covered metal
stents. Ann Thorac Surg 2006;81:467-73.
3. Cheynel N, Arnal E, Peschaud F. et al. Perforation et rupture de l’oeso-
phage: prise en charge et prognostic [French]. Ann Chir 2003;128:
163-6.
4. Freeman RK, Van Woerkom JM, Ascioti AJ. Esophageal stent place-
ment for the treatment of iatrogenic intrathoracic esophageal perfora-
tion. Ann Thorac Surg 2007;83:2003-8.
5. Kiev J, Amendola M, Bouhaider D, et al. A management algorithm for
esophageal perforation. Am J Surg 2007;194:103-6.
6. Koch S, Weber A, Fein F, et al. Esophageal stents as a salvage therapy
for non malignant iatrogenic esophageal perforations. Gastroenterol
Clin Biol 2005;29:735-9.
7. Mason GR. Esophageal perforations, anastomotic leaks, and strictures:
the role of prostheses. Am J Surg 2001;181:195-7.
8. Langer FB, Wenzl E, Prager G, et al. Management of postoperative
esophageal leaks with the polyflex self-expanding covered plastic
stent. Ann Thorac Surg 2005;79:398-404.
9. Siersema PD. Treatment of esophageal perforations and anastomotic
leaks: the endoscopist is stepping into the arena. Gastrointest Endosc
2005;61:897-900.
10. Schubert D, Scheidbach H, Kuhn R, et al. Endoscopic treatment of
thoracic esophageal anastomotic leaks by using silicone-covered self
expanding polyester stents. Gastrointest Endosc 2005;61:891-6.
11. Raju GS, Thompson C, Zwischenberger JB. Emerging endoscopic
options in the management of esophageal leaks. Gastrointest Endosc
2005;62:278-86.
12. Johnson E, Lundell L, Ledman B. Sealing of esophageal perforation or
ruptures with expandable metallic stents: a prospective controlled
study on treatment efficacy and limitations. Dis Esoph 2005;18:262-6.
13. Radecke K, Gerken G, Treichel U. Impact of a self-expanding, plastic
esophageal stent on various esophageal stenoses, fistulas, and leak-
ages: a single-center experience in 39 patients. Gastrointest Endosc
2005;61:812-8.
14. Radecke K, Lang H, Frilling A, et al. Successful sealing of benign esoph-
ageal leaks after temporary placement of a self expanding plastic
stent without fluoroscopic guidance. Z Gastroenterol 2006;44:1031-8.
15. Dormann AJ, Wigginghaus B, Deppe H, et al. Successful treatment of
esophageal perforation with a removable self-expanding plastic stent.
Am J Gastroenterol 2001;96:923-4.
16. Gelbmann CM, Ratiu NL, Rath HC, et al. Use of self-expandable plastic
stents for the treatment of esophageal perforations and symptomatic
anastomotic leaks. Endoscopy 2004;36:695-9.
Received June 10, 2008. Accepted September 18, 2008.
Current affiliations: Departments of Gastroenterology (L.A., C.M., P.A.S.,
C.S., J.-C.G., M.B.) and Digestive Surgery (S.B., O.E., C.B.), Hôpital Nord,
Marseille, France.
Reprint requests: Marc Barthet, MD, PhD, Department of
Gastroenterology, Hopital Nord, Chemin des Bourrely, 13915 Marseille
Cedex 20, France.
- : 2009 www.giejournal.org