Introduction
Endoscopic treatment has been described as a minimally invasive approach for management
of patients with surgically altered anatomy and biliary adverse events (BAEs) with
two critical issues: (1) ability to reach the biliary anastomosis; and (2) performance
of an effective therapeutic procedure. Duodenoscopes, gastroscopes and pediatric colonoscopes
have all been used, with a poor rate of biliary anastomosis achievement [1]. Enteroscopes (short, long, single- or double-balloon) [1]
[2]
[3]
[4]
[5] have also been utilized, with an overall technical success rate of about 76 %. Finally,
a high success rate (around 90 %) has also been described for the EUS-guided approach,
but still with high rates of morbidity (18 %-50 %) and mortality (1.5 %) [6]
[7].
The possibility of performing endoscopic entero-enteral bypass (EEEB) has been explored
in patients with bilio-digestive anastomosis and BAEs [8]
[9]
[10]
[11], but data are scant and with limited follow-up. In 2019, our group performed a study
to retrospectively evaluate the safety, feasibility, and effectiveness of EEEB for
the management of BAEs in 32 patients with Roux-en-Y reconstruction or after a Whipple
procedure. We reported recurrence of BAE in two patients and a cumulative AE rate
of 18.7 % during a mean follow-up of 34.5 ± 23.5 months [12]. We performed this update study to assess the long-term (1–7 years) effectiveness
of EEEB for BAEs.
Patients and methods
Data on all consecutive patients with surgically altered anatomy who underwent EEEB
between January 2014 and January 2021 were retrospectively retrieved from a prospectively
collected database (Niguarda Ethical Committee, n. 43r/2021). Indications for performing
EEEB were the same as in our previous study: (1) inability to endoscopically reach
the biliary site using 140 cm pediatric colonscope; and (2) pre-procedure high probability
of the need for complex biliary procedures defined as presence of multiple strictures
involving both the common bile duct and segmental ducts, (usually related to ischemic
colangiopathy) or presence of multiple duct anastomosis (i. e. biduct bilio-digestive
anastomosis). We do not use enteroscopes in the setting of ERC in altered anatomy
to treat complications of hepaticojejunostomy. Definition and severity of the AEs
were based on the “AGREE” classification by Nass KJ et al. [13]. Exclusion criteria for the present study were surgically altered anatomy without
hepaticojejunostomy, < 1 year follow-up after EEEB, presence of malignant biliary
stricture (excluded by dedicated imaging, i. e. computed tomography and/or magnetic
resonance imaging with contrast) and no indication for EEEB.
To perform the access to the biliary jejunal loop (to recognize it under EUS), four
different types of puncture techniques were used: 1) percutaneous transhepatic biliary
drainage (PTBD) performed before EEEB session; 2) transgastric EUS-guided puncture
of the left hepatic duct; 3) direct EUS-guided puncture of the jejunal loop (recognizing
the anastomotic area both under EUS-guide for the proximity to the hepatic hilum and
under fluoroscopic check for the position of echoendoscope tip); and 4) placement
of a 7F endoscopic tube into the jejunal loop in a retrograde manner.
To create EEEB, Hot Axios Electrocautery and Delivery system (Boston Scientific Corp.)
(15-mm diameter, 10-mm length) both under EUS and fluoroscopic guidance was used.
The use of a Nagi stent in our first experience was abandoned in favor of the Hot
Axios system.
ERC was performed after EEEB during the same session if the axis of the enteral bypass
allowed an in-line approach to the bilio-digestive anastomosis. In all other cases,
the ERC was done 1 week later to allow for anastomosis consolidation [12]. ERC was generally performed using a 3.8-mm operative channel pediatric colonoscope
(ED-127, GS-125, Pentax Europe). Postprocedure management did not change and first
endoscopic follow-up was scheduled 6 months after the index procedure in the setting
of hepaticojejunostomy. Enteral stents for EEEB were generally removed at 12 months
after EEEB creation in asymptomatic patients.
The primary outcome was to evaluate the long-term efficacy of the method defined as
absence of strictures and symptom recurrence at > 1 year follow-up. BAE recurrence
was documented based on new onset of symptoms and new changes in of blood tests for
cholestasis. Magnetic resonance cholangiopancreatography was performed in uncertain
cases. BAE recurrence was finally always confirmed at cholangiography.
Secondary outcomes were: 1) technical success rate defined as the creation of EEEB
and the treatment of BAE through it; 2) clinical success rate defined as resolution
of BAE associated with improvement of clinical status; and 3) AEs related to EEEB
and ERC. AEs were defined as early if occur within 1 week after EEEB.
Partial migration of the EEEB stent was defined as asymmetrical dislodgement of the
flairs on the jejunal or gastric edges close to complete migration.
Complete migration was defined as the absence of the enteral stent at the endoscopic
check.
Quantitative variables were described with mean value and standard deviation. Categorical
values were described with frequencies (%).
Results
During the study period, 80 consecutive patients underwent EEEB (32 patients from
January 2014 to December 2017 plus 48 patients from January 2018 to January 2021).
Our first case series of 32 patients was published in 2019 [12]. A flowchart of the study structure is shown in [Fig. 1]. Four patients were excluded from the present study because they underwent surgery
without creation of hepaticojejunostomy (2 underwent total gastrectomy with Roux-en-Y
reconstruction, 1 bariatric gastric bypass and 1 subtotal gastrectomy with Billroth
II reconstruction). The other eight patients with hepaticojejunostomy underwent EEEB
before March 2022 but were excluded from the present study because of lack of sufficient
follow-up. The other 42 patients with Roux-en-Y reconstruction were treated reaching
the papilla/hepaticojejunostomy directly using a pediatric colonoscope and were excluded
from the present study because they did not undergo EEEB. In the other 106 excluded
patients, complications of biliary anastomosis after a Whipple procedure were treated
retrograde using a pediatric colonoscope. No patient was treated by device-assisted
enteroscopy because of the still low efficacy of treatment and scant long-term clinical
outcomes fully demonstrated in this setting.
Fig. 1 Flowchart of the structure of the present study
The type of surgically altered anatomy is shown in [Fig. 2]. Hepaticojejunostomy on Roux-en-Y loop was the most common type, accounting for
91 % of cases. Indications were biliary single-duct anastomotic stricture in 51 (74 %),
biliary multiple-ducts anastomotic strictures in four (5.7 %), uncovered self-expandable
metal stents entrapment (placed in other hospitals by interventional radiologists)
in four (5.7 %), choledocolithiasis in five (7.2 %), recurrent cholangitis in two
(2.9 %), dehiscence of biliary anastomosis in one case (1.4 %), external biliary fistula
after left hepatectomy in a previous orthotopic liver transplantation (OLT) in one
(1.4 %) and biliary fistula after pancreaticoduodenectomy in one (1.4 %). Six patients
with single-duct anastomotic stricture had undergone orthotopic liver transplantation
(OLT) on Roux-en-Y loop, five of whom had a second liver transplantation after failure
of the first OLT.
Fig. 2 Type of surgically altered anatomy treated in the case series.
In our first series, PTBD was previously positioned in 29 patients, while in other
three patients a 7F catheter was used to allow EEEB creation. Of the new 48 cases,
the puncture techniques were PTBD in 11 patients, transgastric EUS-guided approach
in 16 patients, direct EUS-guided puncture in eight and 7F endoscopic tube in 13 patients.
Overall results of our updated series are presented in [Table 1]. EEEB was successfully created in 79 of the 80 patients (98.7 %), with one failure
due to difficulty in maintaining correct echoendoscope position. A 15-mm and 10-mm
Hot Axios electrocautery system was used in all updated cases.
Table 1
Main results of the series of patients treated with EEEB-ERC.
|
Total number of patients EEEB (2014–2021)
|
80
|
|
Technical success
|
98.7 %
|
|
Clinical success
|
98.7 %
|
|
Recurrence rate
|
3.8 %
|
|
Morbidity
|
32 %
|
|
Mortality
|
0 %
|
|
Mean follow-up (years)
|
4
|
EEEB-ERC, entero-enteral endoscopic bypass endoscopic retrograde cholangiography.
ERC could be performed in all patients, in one session (immediately after EEEB creation)
in 57 patients. In one case, the hepaticojejunostomy was far from the site of EEEB
and not reachable retrogradely using a pediatric colonscope due to strict angulation
of the biliary jejunal loop; in this case, ERC were performed “from a distance” using
standard devices. PTBD was removed immediately after ERC in 36 patients. In the remaining
four cases the PTBD was left in situ to perform bile duct washing in case of multiple
and massive biliary lithiasis associated with the anastomotic stricture (3 patients)
and in the case of the still unreachable anastomosis during the 6-month anastomotic
stenting to have a second way to assist with stent removal (1 case). A total of 14
AEs were reported considering the whole series: seven early and seven late as defined
above. Five early AEs were reported in the new case series (3 grade I, 2 grade III
according to AGREE classification) plus two cases reported in the previous case series
(1 moderate self-bleeding, 1 partial intraprocedural stent displacement): a partial
EEEB stent displacement occurred (where second stent was needed to maintain the anastomosis
opened), two patients developed moderate self-limited bleeding, one patient developed
acute respiratory failure (that was managed with noninvasive ventilation), one case
of asymptomatic intraperitoneal free air under diaphragm during ERC (probably due
to a partial detachment of the enteral loops). The EEEB allowed performance of successful
and effective different types of treatment ([Fig. 3], [Fig. 4], [Fig. 5]).
Fig. 3 Hepaticojejunostomy stricture treated through EEEB. a Endoscopic view of the anastomotic stricture. b Fluoroscopic view of the endoscopic treatment placing two 8.5 F 5-cm plastic stent
(in in the posterior right hepatic duct and one in the left duct) and one 12-mm, 2-cm
fc-SEMS (Nagi stent). c Endoscopic view of the enlarged anastomosis at 6 months, immediately after stent
removal. d Fluoroscopic check after stent removal.
Fig. 4 Management of entrapped uncovered metal stent placed through hepaticojejunostomy.
a Endoscopic view of the stent and foreign body forceps used to remove it. b Endoscopic view of the hepaticojejunostomy immediately after stent removal showing
residual stricture. c Stenting performed using three 8-mm, 3-cm fc-SEMS (Wallflex) to avoid recurrent stricture.
d Fluoroscopic view of the three fc-SEMS positioned through EEEB, e, f Endoscopic view of the hepaticojejunostomy after stents removal 6 months after the
index procedure.
Fig. 5 Hepaticojejunostomy stricture associated with multiple biliary stones treated through
EEEB after percutaneous transhepatic trans-anastomotic biliary drainage. a Fluoroscopic view of EEEB creation in direct line to the anastomosis. b Fluoroscopic view of the treatment positioning two 6-mm, 2-cm fc-SEMS (in the left
hepatic duct and anterior right duct) and one 8-mm, 4-cm fc-SEMS (in the posterior
right hepatic duct) through anastomotic stricture. c Fluoroscopic view at 1 week with biliary stent in site and PTBD left in site for
biliary cleansing using sterile saline. d Fluoroscopic view after stent removal at 6 months showing massive aerobilia in the
whole biliary tree.
In the previous study, four EEEB long-term AEs were reported (grade III according
to AGREE classification): spontaneous EEEB stent displacement, three complete and
one partial, which were solved with insertion of another stent through the stent.
In the new case series, there were two more spontaneous EEEB stent dislodgements,
one complete and one partial, which were resolved with insertion of another stent
as before. We did not report any AEs related to biliary stent placement. In three
patients, considering the whole series, biliary stents had spontaneously migrated
at the time of the second endoscopic check. No differences in timing of enteral stent
removal were recorded (12.7 ± 5.4 months vs12.3 ± 5.5 months) and there were no AEs
related to removal. Seven cases of mucosal pressure ulcers without perforation were
observed at the endoscopic check after removal. No difference in terms of migration
rate was noted if ERC and EEEB were performed during the same session or not.
Mean follow-up of patients was 4 ± 3 years. Two recurrent BAEs occurred in the previous
study, one case of choledocolithiasis and one case of biliary stricture; in the updated
period (2018–2021), one more recurrence of biliary stricture was reported. Recurrences
were successfully retreated through the EEEB. Mean time of recurrence since the first
treatment was 14.3 months (12 and 16 months for stricture recurrence; 15 months for
choledocolithiasis). During the follow-up period, three patients died: one because
of recurrence of previous neoplasm (no cancer recurrence was observed at the level
of hepaticojejunostomy), one because of hepatic insufficiency post-OLT due to ischemic
cholangiopathy and one due to sequalae of ischemic neurological acute accident. No
deaths related to EEEB were reported.
Discussion
After the few successful cases in the first study, we continued to use this technique
for all patients with benign BAEs following Roux-en-Y reconstruction or after a Whipple
procedure, in whom the biliary anastomosis could not be reached endoscopically or
the pre-procedure probability for the need of complex biliary procedures was high.
Indications for the procedure have increased since the previous experience: EEEB also
has successfully been performed in two cases of biliary anastomotic dehiscence in
which both redo surgery and interventional radiology failed to achieve clinical success
because of the coexistence of necrotizing infected tissue in the intestinal loop around
the anastomotic area.
The approach to perform EEEB changed between the two periods in favor of EUS-guided
techniques.
In our updated experience, the transgastric approach and direct puncture of the jejunal
loop both seem to be safe.
Unfortunately, anterograde approach for performing EUS-hepaticogastrostomy alone cannot
overcome the limitations of interventional radiology and still is associated with
high rates of morbidity and mortality [6]
[7].
The transition from Nagi stent placement to Hot Axios system has made the procedure
safer and faster. However, the instability of position with the echoendoscope and
the incorrect opening (just in the middle between the “biliary loop” and the alimentary
loop under fluoroscopy) lead to technical failure and procedural AEs. Moreover, the
previous more challenging experience using Nagi stent has improved the operators’
skills and their confidence in using the Hot Axios system for EEEB creation.
EEEB was successfully performed in all but one patient from the previous study; no
technical failures were experienced using the Hot Axios system. The rate of procedure-related
AEs remained acceptable (7,2 %) and they were related to the creation of the enteral
anastomosis (self-limited bleeding due to unvoluntary puncture of enteral wall vessels,
free intraperitoneal air due to the timing of stent release). The strategy for use
of a large-bore fully-covered SEMSs was successful in terms of recurrence rate: the
jejunal portion of the anastomotic stricture likely is better dilated by SEMS than
multiple plastic stents [12]. The fc-SEMS were “hand-made” tailored in the present study because the length of
the commercially available stents was too long to avoid obstruction of the hepatic
hilum on one side or mucosal ulcer on the jejunal edge. The stent was cut with sterile
scissors once freed from the delivery system. At the end, it was reinserted in the
delivery system ready to be placed at the site of the anastomosis. This approach was
necessary while improvement in the commercially available stents is awaited.
The recurrence rate remained low in the extension of follow-up. Gradual and long-lasting
dilation of the stricture produced by SEMSs likely is effective in achieving sufficient
long-term control for fibrotic healing of the anastomosis. This is probably related
to the fact that the bilio-digestive anastomotic stricture is a compound stenosis
(half biliary component and half enteral).
If the biliary part also can be correctly dilated pneumatically, the enteral edge
does not respond properly to this type of dilation, as happened with a pure enteral
stricture. This is one of the main reasons for scant clinical results from long-term
follow-up of interventional radiology in this setting [14], as we discussed in our previous paper [12]. Considering all the cases in which a PTBD was present before EEEB creation, all
40 patients met the inclusion criteria and had anastomotic stricture. In all the cases,
PTBD did not provide complete drainage of the biliary tree and only two possible therapeutic
chances were proposed: 1) iterative pneumatic dilation of the anastomosis in case
of short common bile duct; and 2) placement of biliary percutaneous stents. We have
already discussed the long-term effects of pneumatic dilation in this setting, which
is associated with a high percentage of major complications [15]. Furthermore, placement of percutaneous stents also has the limitation of use of
long stents, which are associated with development of hyperplastic tissue and secondary
sludge on the biliary side and risk of jejunal decubitus, perforation and higher risk
of migration on the enteral side of the anastomosis. In comparison with EEEB, the
percutaneous approach requires many punctures to treat multiple anastomotic strictures
and percutaneous stent removal may be more difficult than an endoscopic one. Moreover,
in case of entrapped biliary stents, the possibility of managing this AE on the enteral
edge led to better treatment due to shorter flogistic ingrowth on this side than on
the biliary one.
Long-term EEEB patency has been confirmed in the present study. Spontaneous stent
migration is the main long-term AE of EEEB. In the majority of cases, the entero-enteral
bypass remained patent; in a few cases, partial migration can lead to one flair of
the stent being buried, usually through the gastric wall (due to flogistic reactions).
In the latter cases, we were able to save the endoscopic anastomosis by placing another
stent within the previous one [16]. Pressure mucosal ulcers at the site of the flairs at the time of enteral stent
removal did not always require any further treatment. Considering the AEs in their
entirety, they should be considered mild because early ones were resolved during the
same endoscopic session and did not require further procedures and grade III AEs were
stratified in this manner only because an additional endoscopic procedure was needed.
Limitations of the study are its retrospective nature, which does not eliminate the
chance of bias; the heterogeneity of the treated patients; and the single-center nature
of the study, which does not allow for reproducibility of the procedure.
Conclusions
In conclusion, the present update study confirmed that EEEB in patients with hepaticojejunostomy
is safe, feasible, and allows for very effective treatment of different BAEs, based
on more than 7 years of experience and research in this field.
