Introduction
Plastic stent migration after endoscopic biliary drainage can happen in approximately
5 % of cases and may very rarely lead to lateral duodenal wall perforation [1]
[2]. Treatment options when this occurs include endoscopic treatment with through-the-scope
(TTS) clips or over-the-scope clips (OTSC) and surgical repair [3]
[4]. The latest option is recommended when perforation is not diagnosed immediately
(> 12 h) and when there is contrast extravasation or intra-abdominal fluid collection
[4]. Successful endoscopic treatment of these perforations using OTSC, even when it
was diagnosed more than 12 hours after the initial insult, was reported in a few case
reports [5]
[6]
[7], but data about efficacy of such treatment in this indication remain sparse. We
aimed to review and describe the clinical characteristics and outcomes of all cases
of lateral duodenal wall perforation due to migrated biliary stent that were endoscopically
treated using OTSC in a tertiary referral center over 50 months.
Case reports
Cases were obtained by systematic review of the prospective database including all
endoscopic retrograde cholangiopancreatographies (ERCPs) performed in our center from
January 1, 2014 to March 31, 2019 identifying all cases having duodenal perforation
secondary to biliary plastic stents displacement. Individual data of the screened
cases including demographics, stenting indication, migrated stents’ characteristics,
time from stent placement to perforation, time from perforation diagnosis to endoscopic
treatment and finally immediate and 28-day treatment outcome were extracted.
Over this period, out of 696 ERCPs in which plastic stents were placed, 6 cases (0.8 %)
of lateral duodenal wall perforation ([Fig. 1]) caused by plastic biliary stent displacement were identified. Their main characteristics
are presented in [Table 1]. In most cases (5/6) the diagnosis of perforation was established early after the
stent placement (< 5 days) but none was made in less than 12 hours. All five patients
developed abdominal pain and fever and the diagnosis was made based on the computed
tomography (CT) imaging. The other patient had no symptoms and a stent perforating
the duodenal wall was found during an elective ERCP planned for stents replacement
90 days after the initial one.
Fig. 1 Radiologic and endoscopic pictures from patient 6. a CT-scan showing migration through the duodenal wall of the stent placed in the left
hepatic ducts. b Endoscopic image of the double flap plastic stent perforating the duodenum. c Lateral duodenal wall perforation as seen after stent removal.
Table 1
Patient characteristics.
|
Gender, age
|
Stenting indication
|
Location of the stricture/leak
|
Days between stent placement and perforation diagnosis
|
Bray classification [1]
|
Stents placed/ migrated (N/n)
|
Culprit stent technical characteristics
|
Immediate outcome
|
Continued stenting during the same procedure?
|
28 days outcome
|
|
1. M 75
|
Biliary leak post hepatectomy
|
Hilar leak
|
4
|
I
|
2/1
|
Double flaps
18 cm 8.5Fr
|
Success at closure
|
No 6 Fr Nasobiliary catheter
|
No further interventions or complications
|
|
2. M 61
|
Ischemic cholangiopathy (BBS) and biliary leak post liver transplantation
|
Anastomotic leak
|
2
|
II
|
1/1
|
Sigmoid-shaped
17 cm 8.5Fr
|
Success at closure
|
No 6 Fr Nasobiliary catheter
|
No further interventions but death 17 days later (biliary sepsis)
|
|
3. F 31
|
Choledocholithiasis and bile duct stenosis (BBS)
|
Common bile duct stenosis at the junction with cystic duct
|
4
|
IV
|
2/1
|
Double flaps
15 cm 7Fr
|
Success at closure
|
No
|
No further interventions or complications
|
|
4. M 52
|
Ischemic cholangiopathy (BBS)
|
Sub-hilar stenosis
|
90
|
I
|
3/2
|
Double flaps
12 cm 8.5Fr
|
Success at closure
|
Yes Double flaps
12 cm 8.5 Fr RIHD
7 cm 10 Fr X 2 CBD
|
No further interventions or complications
|
|
5. M 72
|
Bile duct compression after hepatic artery embolization (BBS)
|
Proximal common bile duct stricture
|
2
|
IV
|
2/1
|
Double flaps
13 cm 8.5Fr
|
Success at closure
|
Yes Double flaps
10 cm 10 Fr X2 RIHD
12 cm 10 Fr X1 RIHD
|
Peritonitis needing laparotomy and death at day 5 post perforation
|
|
6. F 45
|
Anastomotic stenosis post liver transplantation (BBS)
|
Anastomotic stricture
|
2
|
IV
|
2/1
|
Double flaps
12 cm 8.5Fr
|
Success at closure
|
Yes Double flaps
10 cm 10 Fr RIHD
10 cm 10 Fr LIHD
|
No further interventions or complications
|
BBS, benign biliary stricture; RIHD, right intrahepatic ducts; LIHD, left intrahepatic
ducts; CBD, common bile duct
Regarding indication for biliary stenting, five patients were treated for benign biliary
stricture (BBS) (three after liver transplantation, one secondary to multiple common
bile duct stones and one due to external compression after hepatic artery embolization).
The remaining patient had a biliary leak after undergoing right hepatectomy for colorectal
cancer metastasis.
Certain similarities are evident regarding the technical characteristics of the stents
used among the reported cases ([Table 1]). First, for most of the patients (5/6), biliary stenting was done with either one
or two stents (one and four patients, respectively). Inversely, only one patient had
multiple (≥ 3) stents in place. Double flaps plastic stents were used in four of them
and in one patient a single preformed sigmoid-shaped plastic stent was used.
In all cases the longest stent was the culprit and in five of six patients it migrated
from the left intra-hepatic bile ducts. The length of the perforating stents ranged
from 12 cm to 18 cm. Moreover, two patients with BBS had a history of previous stenting
achieving a higher calibration diameter than what was applied at the last examination.
More specifically, Patient 3 had a previous maximum calibration of 18.5 Fr while the
total diameter of stents in his last ERCP was of 10 Fr and Patient 4 had a maximum
calibration of 27 Fr while the total diameter of stents in his last ERCP was 25.5
Fr.
In all cases, the perforating stent was removed with rat tooth forceps after aspiration
of digestive fluid, using a therapeutic gastroscope (Olympus GIF-1T160; 1T190) under
CO2 insufflation. Immediately thereafter, a T-type teeth 12-mm over-the-scope clip
(OTSC) (Ovesco Endoscopy AG, Tübingen, Germany) was placed on the scope and advanced
facing the perforation. In every patient, perforation was the size of a single plastic
stent (average diameter 3 mm) and there was no fibrotic tissue on the edges of the
perforation. The sides of the perforation were grasped using the twin grasper (Ovesco
Endoscopy AG) and pulled back into the hood with low aspiration before releasing the
clip. Duodenal contrast opacification was performed during the same session and, if
needed, new biliary stents were inserted afterwards. Broad-spectrum antibiotics were
started from the perforation diagnosis and were given for a median time of 9.5 days
(3–17 days). A nasogastric tube was inserted in three of six cases at the end of the
endoscopic procedure, however, we think that it probably wasn’t necessary because
immediate contrast injection after OTSC application did not show any residual leak
in all procedures. Successful closure of all perforations was confirmed by subsequent
imaging modalities ([Fig. 2]).
Fig. 2 Radiologic and endoscopic pictures from patient 2. a Over-the-scope-clip mounted on the tip of the scope before application; b Fluoroscopic image obtained right after placement of OTSC. c Contrast injection showing no leak after the application of OTSC
In five patients no further interventions were needed and outcome at 28 days was excellent
for four of them. After initial improvement of his medical condition, one patient
(Patient 2) died at day 17 after perforation closure. His death was attributed to
septic shock secondary to cholangitis. Further intervention after successful closure
of the perforation was needed for only one patient who developed peritonitis secondary
to the perforation and required laparotomy and peritoneal lavage. The laparotomy did
not show any residual leak at the perforation site but he eventually died of this
complication at day 5 post procedure. This patient was an elderly and malnourished
man with important comorbid vascular disease. He also had a more worrying CT at time
of diagnosis with a large quantity of peritoneal and retroperitoneal air as well as
ascites ([Fig. 3])
Fig. 3 Computer tomography at the time of the diagnosis showing an important retropneumoperitoneum
(Patient 5)
Discussion
As far as we know, our case series is the first to describe in a systematic way the
incidence and treatment of duodenal wall perforation by plastic biliary stent migration
using OTSC. The main finding is that wall closure was obtained in 100 % of cases,
but this is probably not enough in some cases since clinical outcome obviously depends
on the severity of the peritonitis and on the general state of the patient.
In our tertiary referral center, six cases of lateral duodenal wall due to plastic
stent displacement were identified over 50 months (< 1 % of overall ERCPs with biliary
plastic stenting; a number similar to the one reported in the literature [2]). Risk factors associated with distal stent migration are not well studied. Potential
ones include BBS (against malignant biliary strictures), stent shape as well as single
stenting [8]
[9]. These data correlate with our case series. All but one patient with distal migration
of a plastic biliary stent had a BBS, and the last one had a postsurgical leak. Of
note, half of them were associated to benign stricture due to previous liver transplantation.
Moreover, single or double stenting was the most common practice in our cases. Only
one patient who presented with distal migration was previously treated with placement
of three plastic biliary stents. Regarding the length of the stents, all lateral duodenal
wall perforations in our cohort were associated with migration of long stents (≥ 12 cm),
often used in the setting of anastomotic stenosis post-orthotopic liver transplantation
or of perihilar stenosis. Similarly, longer stents (≥ 13 cm) were more susceptible
to migrate distally in a retrospective study of more than 520 procedures [8]. Temporary placement of multiple plastic stents with repeated interventions and
calibration is recommended for the treatment of BBS [10]. Interestingly, in our cohort, two patients had a history of higher maximal calibration
of a BBS than the one achieved at the last stenting before perforation. Considering
this, we adopted and suggest the policy to always replace biliary stents by at least
as many as the number in place before the procedure, especially when the stents are
long or bypass the hilum.
OTSC appears to be an effective treatment with a technical success rate of 100 % in
our cohort without any immediate complications. In all patients, follow-up images
confirmed that the breach was closed ([Fig. 2]). OTSC offers some potential advantages over TTS clips that have also been used
to treat these perforations [3]. The main one is that OTSC allows a larger area of tissue to be grasped, a feature
potentially beneficial in these stent-induced perforations in which the size at the
mucosal level does not necessarily reflect the extent of damage at the layers beyond
mucosa. As shown in our case series, OTSC allowed successful sealing of all duodenal
wall layers at one time.
Finally, in our series, four of six patients treated with OTSC had a good outcome
even if the perforation was diagnosed later than 12 hours after it happened. However,
two patients died, including one from complications directly related to the perforation.
This highlights the need to select patients who can be treated endoscopically but
also to identify early those who could benefit from additional percutaneous drainage
or surgery with peritoneal lavage. Patients with a more severe presentation (clinical
or radiological) may be the ones that would require a more invasive treatment. Reviewing
each patient’s CT scan at the time of diagnosis, the patient who did need further
intervention (Patient 5) had severe retropneumoperitoneum and significant ascites
(Bray type IV perforation) [1]. However, this CT severity index is not perfect since two other patients with Bray
type IV perforation had a good clinical evolution.
Conclusion
In conclusion, lateral duodenal wall perforation by biliary stent migration is a rare
complication that was mostly treated surgically. Even if there are obvious limitations
in this study because of the retrospective design and the low number of patients,
this case series suggests that endoscopic treatment with OTSC can be successful without
a subsequent invasive procedure in selected patients and should be incorporated into
the treatment algorithm of this complication. More data are needed to identify which
patients could benefit from endoscopic treatment and which patients would be better
served having more invasive treatment with surgery.
