ARTICLE IN PRESS

Surgery ■■ (2018) ■■–■■

Contents lists available at ScienceDirect

Surgery
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y m s y

Long-term follow-up and risk factors for strictures after

hepaticojejunostomy for bile duct injury: An analysis of surgical and
percutaneous treatment in a tertiary center
Klaske A.C. Booij a, Robert J. Coelen a, Philip R. de Reuver a,b, Marc G. Besselink a,
Otto M. van Delden c, Erik A. Rauws d, Olivier R. Busch a, Thomas M. van Gulik a, and
Dirk J. Gouma a,*
a Department of Surgery, Academic Medical Center, Amsterdam, The Netherlands
b Department of Surgery, Radboud University, Nijmegen, The Netherlands
c
Department of Radiology, Academic Medical Center, Amsterdam, The Netherlands
d Department of Gastroenterology, Academic Medical Center, Amsterdam, The Netherlands

A R T I C L E I N F O A B S T R A C T

Article history: Background. Hepaticojejunostomy is commonly indicated for major bile duct injury after cholecystec-
Accepted 3 January 2018 tomy. The debate about the timing of hepaticojejunostomy for bile duct injury persists since data on
postoperative outcomes, including postoperative strictures, are lacking. The aim of this study was to analyze
short- and long-term outcomes of hepaticojejunostomy for bile duct injury, including risk factors for
strictures.
Method. Analysis of outcome of hepaticojejunostomy in bile duct injury patients referred to a multi-
disciplinary team.
Results. Between the years1991 and 2016, 281 patients underwent hepaticojejunostomy for bile duct
injury. Clavien-Dindo grade III complications occurred in 31 patients (11%) and 90-day mortality oc-
curred in 2 patients (0.7%). After a median follow-up of 10.5 years (interquartile range 6.7–14.8 years),
clinically relevant strictures were found in 37 patients (13.2%). Strictures were treated with percutane-
ous dilatation in 33 patients (89.2%), and 4 patients (1.4%) were reoperated. The stricture rate in patients
undergoing hepaticojejunostomy <14 days, between 14–90 days, and >90 days after bile duct injury was
15.8%, 18.7%, and 9.9%, respectively. The stricture rate for early versus intermediate and late repair did
not differ (P = 0.766 and 0.431, respectively). The stricture rate for repair after 14–90 days, however, was
higher compared with repair >90 days after bile duct injury (P = 0.045). In multivariable analysis male
gender was the only independent variable associated with stricture formation (OR 6.7, 95% CI 1.8–25.4,
P = 0.005).
Conclusion. Hepaticojejunostomy is a relatively safe treatment of bile duct injury. Timing of surgery
and intermediate repair affect long-term stricture rate; most anastomotic strictures can be treated suc-
cessfully with percutaneous dilation.
© 2018 Elsevier Inc. All rights reserved.

Introduction The impact of timing of HJ on postoperative complications and

Bile duct injury (BDI) remains a potentially devastating compli-
cation of laparoscopic cholecystectomy. BDI is associated with
significant morbidity, high costs, impaired quality of life, and de-
creased survival.1-5 After major BDI, reconstructive surgery by
hepaticojejunostomy (HJ) is usually indicated.6-8
* Corresponding author. Department of Surgery, Academic Medical Center,
Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands.
E-mail address: d.j.gouma@amc.uva.nl (D.J. Gouma).
long-term outcomes such as stricture rate are still debated.9 Early
repair may avoid readmissions, improve quality of life, and reduce
costs, but the presence of (local) inflammation, biliary peritonitis,
and sepsis are associated with worse short- and long-term out-
comes and therefore considered as relative contraindications for early
repair.9-11 In late repair, reconstruction is performed after sepsis
control, with less local inflammation and restored bile duct vascu-
larization, which theoretically would lead to better anastomotic
patency.9,10 In a previous study by our center, early repair was defined
as taking place within 6 weeks after the injury and as being delayed
performing the repair after 6 weeks, but this seems not to be a useful
definition, considering the general clinical course after BDI.11 The

https://doi.org/10.1016/j.surg.2018.01.003
0039-6060/© 2018 Elsevier Inc. All rights reserved.

Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
 ARTICLE IN PRESS
2 K.A.C. Booij et al. / Surgery ■■ (2018) ■■–■■
current definition for early repair “within 2 weeks after the injury”
seems more appropriate.12
Literature concerning risk factors for late postoperative stric-
ture formation and subsequent treatment is limited. Percutaneous
transhepatic dilatation and early surgical revision of the anasto-
mosis are scarcely described. The aim of this study is to analyze the
short-term and long-term outcomes of HJ in patients with BDI, to
determine risk factors for stricture formation, and to assess the out-
comes of subsequent treatment.
Methods
Design and patients
We reviewed data from a prospective cohort of consecutive pa-
tients with BDI, who were referred to the Academic Medical Center
in Amsterdam between the years 1991 and 2016 and treated by HJ
as final treatment. Patients were treated by a multidisciplinary team
of gastroenterologists, (interventional) radiologists, and surgeons.
Because of the retrospective nature of this study, ethical approval
was waived by our institutional medical ethical review board.
Data sources and setting
Clinical data obtained from medical charts and operative reports
from the referring hospital and our center were reviewed. Collect-
ed data included: demographics, indication for cholecystectomy, type
of operation (laparoscopic, conversion, open surgery), American As-
sociation of Anesthesiologists classification, injury classification,13-15
time from injury to diagnosis, interventions in referring hospital,
time from injury to referral, and time from injury to reconstruction.
Treatment and follow-up
Endoscopic, radiological and surgical interventions were per-
formed as previously described.11,16,17
Surgical reconstruction was performed using a Roux-en-Y
hepaticojejunostomy. Dissection was started toward the liver hilum.
The bile duct remnant was identified, if necessary by division of the
hilar plate, as described by Couinaud and Bismuth15 and recom-
mended by Blumgart.18 In some patients, a percutaneous transhepatic
catheter (10-F polyethylene) had been placed during the first hos-
pital admittance for the management of persisting bile leakage or
drainage of bile after a complete occlusion of the bile duct. This in-
dwelling catheter could also be used for identification of the
damaged duct in the hilum. The common hepatic, left and right
hepatic ducts were further dissected and opened, in particular after
the left duct leading to the segment 2 and 3 ducts. The segmental
ducts were sutured together, if possible, to enable the construc-
tion of one, rather than 2 jejunal anastomoses. The transhepatic
catheter, when inserted before surgery, was left behind in the
jejunum loop across the biliary anastomosis and removed after 2–6
weeks, depending on the clinical course, the level of the anasto-
mosis, and the surgeons’ preference. In all patients, the transhepatic
catheter was removed within 6 weeks after surgery. In our center,
high injuries (segmental injury, at the level of the hepatic duct and
confluence) were treated with interrupted sutures (especially the
anterior part of the duct to ensure duct patency). In lower, less
complex injuries, a running suture was used. No technical modifi-
cations had occurred during the study period. The number of
anastomoses was not further analyzed or specified.
Definition of timing of HJ was adapted11 in the present study and
defined as early (<14 days), intermediate (14–90 days) and late (>90
days after BDI). Outcome after reconstructive surgery was ana-
lyzed by the assessment of short-term (<30 days) complications
according to the Clavien-Dindo classification. 19 Short-term
Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
complications evaluated were sepsis, bleeding, intra-abdominal
abscess, liver abscess, cholangitis, cardiopulmonary events, anas-
tomotic leakage, and relaparotomy. The following long-term
complications (>30 days after HJ) were assessed: stricture forma-
tion, incisional hernia, and late reoperation after HJ. Incisional hernia
was defined as a palpable defect in the aponeurosis. BDI-related mor-
tality was assessed as within 30 days, 90 days, and during follow-
up after 90 days after HJ. Patients with a clinically relevant HJ
stricture were defined as those who presented with abdominal
symptoms, cholangitis and/or abnormal liver function tests in con-
junction with a stenosis at the HJ, diagnosed with percutaneous
transhepatic cholangiography (PTC) or on magnetic resonance
imaging, requiring intervention.
Percutaneous transhepatic dilatation was performed during a
period of 3 months in which a patient underwent 3–4 sequential
dilatation procedures. A percutaneous transhepatic dilatation was
defined as successful if no residual significant stricture was seen
on cholangiography and no clinical signs of stenosis were ob-
served after removal of the drainage catheter.17
Recurrent strictures were defined as strictures at the HJ that were
diagnosed either after prior successful treatment of the initial stric-
ture by percutaneous transhepatic dilatation or surgical revision of
HJ. Follow-up data were obtained through regular outpatient visits,
and long-term outcome was obtained by mail and telephone surveys
to the general practitioner and the referring surgeons.
Statistical analysis
All baseline characteristics are represented as numbers and per-
centages. Mean and standard deviation are given for normally
distributed data and median and interquartile range (IQR) for non-
normally distributed data. Chi-square or the Fisher exact tests were
used to compare dichotomized variables between groups of pa-
tients with different timing of repair. For continuous variables, the
Kruskal-Wallis H test was used to calculate differences between
groups for non-normally distributed data. Patient characteristics,
referral and injury characteristics, and timing of surgery were factors
analysed to determine the risk for stricture formation. Indepen-
dent prognostic factors for stricture formation were identified, using
multivariable binary logistic regression analysis including prog-
nostic factors with a P value of < 0.1 in the univariable analysis.
Kaplan-Meier estimates of stricture free survival were obtained.
Stricture-free survival was compared among 3 groups with differ-
ent timing of repair, using log-rank test statistics.
A P value of ≤ 0.05 was considered statistically significant. All sta-
tistical analyses were performed using IBM SPSS Statistics for
Windows v 23.0 (IBM Corp., Armonk, NY, USA).
Results
Patient characteristics
Between the years 1991 and 2016, 281 patients underwent HJ
for BDI in our center. Patients’ demographics are presented in Table 1.
The date of last follow-up was recorded in September 2016. Most
patients were women (67.3%) and the median age was 46.3 years
(IQR 36.5–62.3 years). Most patients were diagnosed with Amster-
dam classification type D (n = 225, 80.1%), Strasberg type E injury
(n = 238, 84.7%). Location of injury was mostly classified as Bismuth
type III injury, at the hepatic confluence (n = 102, 36.3%).
Referral pattern
Patients were referred from nonteaching hospitals, (n = 133,
47.3%), teaching hospitals (n = 115, 40.9%), and academic centers
(n = 5, 1.8%), including 2 patients from our own center (0.7%). Time
 ARTICLE IN PRESS
K.A.C. Booij et al. / Surgery ■■ (2018) ■■–■■ 3

interval of diagnosis of BDI, referral pattern, and interventions before drainage was performed in 36 patients (12.8%) and exploration alone
referral are summarized in Table 1. BDI was diagnosed after a median in 13 patients (4.6%).
of 6 days after laparoscopic cholecystectomy (LC) (IQR 1.0–16.0 days). Patients were referred after a median of 21 days (IQR 9.0–
Most injuries were diagnosed either after LC during admission or 111.0 days) of which 100 patients (35.6%) within 2 weeks after the
after discharge (n = 99, 35.2%, and n = 99, 35.2%, respectively). initial BDI.
Before referral, 69 patients (24.6%) had undergone a surgical
reintervention. In 20 of these 69 patients (29.0%), biliary recon- Short-term and long-term outcomes
struction was performed: HJ in 11 patients, end-to-end repair in 9
patients. In 1 patient, an acute partial liver resection without HJ was Short-term outcomes
performed in the referring hospital because of vascular injury of the Clavien-Dindo grade III complications were observed in 31 pa-
right hepatic artery and right portal vein. This patient was re- tients (11%) (Table 2). In 18 patients (6.4%) biliary leakage occurred
ferred to our center for biliary reconstruction thereafter. Surgical of whom 15 patients (83.3%) were treated by percutaneous ab-
dominal drainage (n = 9), percutaneous transhepatic cholangiography
drainage (PTCD) (n = 2), or a combination of both (n = 2). In 2 other
patients the abdominal drains, inserted during surgery, were left
Table 1
Patient characteristics, moment of diagnosis, referral pattern, and interventions of in situ for 2 and 4 weeks, respectively, and removed after normal
patients with bile duct injury before referral. cholangiography.
Variable N = 281 %, SD, or IQR
Reoperation after HJ was indicated in 5 patients (1.8%) because
of persistent bile leakage requiring surgical drainage (n = 3), sepsis
Age (years, median, SD) 46.3 (36.5–62.3)
attributable to an intra-abdominal abscess (n = 1), or an uncon-
Gender (female) 189 67.3
Indication for cholecystectomy trolled bleed (n = 1). In the patient requiring a reoperation because
Symptomatic cholecystolithiasis 230 81.6 of an intra-abdominal abscess, the HJ was performed in the year
Acute cholecystitis 43 15.2 1992 when, in case of severe sepsis, a relaparotomy was per-
Unknown 8 2.9
formed rather than attempting percutaneous drainage. In the patient
Initial procedure
OC* 36 12.8
with an uncontrolled bleed, a relaparotomy was performed 3 times
LC† (conversion) 245 (93) 87.2 (33.1) without finding a clear bleeding focus. Reoperation was per-
ASA classification‡ formed after a median of 7 days (IQR 2.5–107.5 days).
ASA 1/2 254 90.4 Early mortality (<30 days) occurred in 1 patient (0.4%) because
ASA 3/4 27 9.6
of a cardiac arrest 16 days after the repair. The cause of death was
Type of injury
Amsterdam classification sepsis and cardiac failure. Autopsy showed an adequate anastomo-
B 24 8.5 sis, without signs of bile leakage, bleeding, or cholangitis. In 2
C 32 11.4 patients (0.7%) 90-day mortality occurred.
D 225 80.1
Strasberg classification
B 5 1.8 Long-term outcomes
C 16 5.7 Median follow-up is 10.5 years (IQR 6.7–14.8 years) after HJ. The
D 22 7.8 most frequent long-term complications (after 30 days) were de-
E 238 84.7
velopment of a postoperative stricture at the HJ (n = 37, 13.2%) or
Level of injury
Bismuth classification
incisional hernia (n = 21, 7.5%). Incisional hernia was surgically cor-
Bismuth I 5 1.8 rected in 16 patients (5.7%).
Bismuth II 89 31.7 Late BDI-related mortality (>30 days) during follow-up oc-
Bismuth III 102 36.3 curred in 5 patients: attributed to secondary liver cirrhosis (n = 1),
Bismuth IV 51 18.1
Bismuth V 31 11.0
Unclassified 3 1.1
Moment of diagnosis Table 2
During cholecystectomy 73 26.0 Outcome of hepaticojejunostomy in 281 patients with bile duct injury.
In hospital 99 35.2
N = 281 %, IQR
After discharge 99 35.2
Unknown 10 3.6 Short-term complications (<30 days)
Interventions in referring hospital Sepsis 6 2.1
Surgery Bleeding 13 4.6
Initial repair 50 17.8 Intraabdominal abscess 12 4.3
Direct repair/E-E anastomosis 38 13.5 Liver abscess 9 3.2
HJ/partial liverresection 11/1 3.9/0.4 Cholangitis 48 17.1
Relaparotomy 69 24.6 Cardiopulmonary 13 4.6
With reconstruction (anatomic 20 (9/11) 29.0 (13.0/15.9) Anastomotic leakage 18 6.4
repair/HJ) Early reoperation after HJ <30 days 5 1.8
Exploration (surgical drainage) 49 (36) 17.4 (12.8) Mortality
Radiology <30 days 1 0.4
Percutaneous drainage 78 27.8 <90 days 2 0.7
PTCD§ 28 10.0 Clavien-Dindo classification
Endoscopy Grade I/II 21/29 7.5/10.3
ERCP papillotomy (+stent) 88 (43) 31.3 (15.3) Grade III/IV 31/5 11.0/1.8
Time from injury to diagnosis (days), 6.0 (1.0–16.0) Grade V 1 0.4
median (IQR) Long-term complications (>30 days)
Time from injury to referral (days), 21.0 (9.0–111.0) Postoperative stricture of HJ 37 13.2
median (IQR) Incisional hernia (surgical correction) 21 (16) 7.5 (5.7)
Late reoperation after HJ 4 1.4
* Open cholecystectomy.
† Laparoscopic cholecystectomy. Mortality
‡ >30 days 5 1.8
The American Society of Anesthesiologists physical status classification.
§ Percutaneous transhepatic cholangiography drainage. >90 days 4 1.4

Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
 ARTICLE IN PRESS
4 K.A.C. Booij et al. / Surgery ■■ (2018) ■■–■■

Table 3
Analysis of timing of HJ on postoperative complications.

Early repair Intermediate repair Late repair P value

(<14 days) (14–90 days) (>90 days)

N = 19 N = 91 N = 171

N % N % N %

Leakage HJ 3 15.8 7 7.7 8 2.3 .143

Relaparotomy tertiary center because of complications after HJ 2 10.5 3 3.3 4 1.2 .804
Clavien-Dindo classification >3 5 26.3 12 13.2 20 11.7 .464
Stricture HJ 3 15.8 17 18.7 17 9.9 .129
Early vs Intermediate .766
Early vs Late .431
Intermediate vs Late .045

Chi square test.

persistent abdominal sepsis (n = 3), and ischemic bowel because of
an acute internal herniation (n = 1). Mortality after 90 days oc-
curred in 4 patients (1.4%).
Overall BDI-related mortality occurred in 6 patients (2.1%) after
a median of 44.6 months (IQR 2.4–99.5 months).
Timing of HJ versus complications
In Table 3, the impact of timing of HJ on postoperative compli-
cations is summarized. Early, intermediate, or late repair was not
associated with significant differences in anastomotic leakage,
relaparotomy, or grade III complications. Stricture formation was
present in 3 patients after early (<14 days), 17 patients (18.7%) after
intermediate (14–90 days), and 17 patients (9.9%) after late repair
(>90 days). Stricture formation did not differ between early versus
intermediate and late reconstruction (P = 0.766 and .431, respec-
tively). However, a higher stricture rate was observed in patients
who underwent intermediate versus late repair (18.7% versus 9.9%,
P = 0.045), independent of sepsis before reconstruction (11.1% versus
7.7% respectively, P = 0.565). Kaplan-Meier estimates of stricture-
free survival showed no significant differences among the 3 groups
with different timing of repair (P = 0.12) (Fig 1).
Risk factors for HJ stricture
A clinically relevant HJ stricture was diagnosed in 37 patients
(13.2%), with a median of 30.4 months (IQR 9.5–51.7 months)
postoperatively.
Univariable and multivariable analysis for stricture develop-
ment is summarized in Table 4. In univariable analysis, male gender,
relaparotomy with reconstruction in the referring hospital, sepsis
before reconstruction, Clavien-Dindo complications > grade III, sepsis
in tertiary center, and postoperative bile leakage were associated
with occurrence of HJ stricture. In multivariable analysis male gender
was the only independent variable associated with stricture for-
mation (OR 6.720, 95% CI 1.778–25.400, P = 0.005). No differences
were found for the timing of reconstruction.
HJs were performed between the years 1991 and 2014. Our early
period was defined as before the year 2003 (n = 95) and later period
as HJs performed after the year 2003 (n = 186). We found no sig-
nificant difference in stricture formation between these 2 time
periods (n = 15/95, 15.8% before 2003 versus 22/186, 11.8% after
2003; P = 0.353). The median repair for HJ before 2003 was 116 days
(IQR 69–283) and after 2003 was 113 days after LC (IQR 71–391
days).
During these 2 periods the incidence of acute repair (<2 weeks)
was not significantly different (n = 8/95, 8.4% before 2003 versus
n = 11/175, 6.3% after 2003) (P = 0.428).
Percutaneous treatment of HJ strictures
Fig 2 illustrates the success or failure with subsequent treat-
ment after initial PTC and dilatation. Only 1 patient (0.4%) underwent
primary surgical revision of the HJ stricture. This was in 1994, in a
period in which we had limited access to PTCD treatment and ad-
equate dilatation procedures. During follow-up this patient developed
a recurrent stricture that was initially treated by dilatation and finally
by secondary revision of the HJ.
In 33 patients (89.2%), the initial treatment of the HJ stricture
was by percutaneous dilatation. In total, 19 of the 33 patients (57.6%)

Fig. 1. Kaplan-Meier analysis of stricture formation after reconstructive surgery for bile duct injury in acute (<14 days), intermediate (14–90 days), and late (>90 days) phase.

Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
 ARTICLE IN PRESS
K.A.C. Booij et al. / Surgery ■■ (2018) ■■–■■ 5

Table 4
Risk factors for stricture formation after HJ in a tertiary referral center.

Univariable analysis Multivariable analysis

OR (95% CI) P value OR (95% CI) P value

Patient characteristics
Age (n = 281) 1.012 (0.990–1.034) .306
Gender male (n = 92) 3.198 (1.578–6.483) .001 6.720 (1.778–25.400) .005
ASA (n = 281) 1.287 (0.767–2.160) .340
Type of injury
Strasberg classification
E 1.000 (reference)
D 1.5 (0.163–13.827) .721
C 0.400 (0.051–3.128) .383
B 0.286 (0.037–2.194) .228
Amsterdam classification 1.000 (reference)
Type D
Type C 0.246 (0.310–1.805) .164
Type B 0.175 (0.230–1.325) .092
Bismuth classification 1.382 (0.636–3.003) .413
I/II vs III t/m V
(n = 94 vs n = 184)
Treatment before referral
Initial repair referring hospital (n = 50) 0.880 (0.346–2.237) .788
Relap reconstruction referring hospital (n = 20) 3.180 (1.138–8.882) .027 1.411 (0.109–18.207) .792
In tertiary center
Time injury to referral (n = 255) 1.000 (0.999–1.000) .759
Sepsis before reconstruction (n = 24) 3.061 (1.173–7.987) .022 0.844 (0.076–9.396) .890
Time of repair
Early (n = 19) 1.000 (reference)
Intermediate (n = 91) 1.225 (0.320–4.684) .767
Late (n = 171) 0.589 (0.156–2.228) .435
Within 6 weeks (n = 25) 1.000 (reference)
After 6 weeks (n = 256) 0.571 (0.200–1.629) .571
Clavien-Dindo > 3 (n = 37) 3.808 (1.191–12.173) .024 3.690 (0.820–16.614) .089
Sepsis tertiary center (n = 6) 7.088 (1.375–36.544) .019 3.192 (0.266–38.327) .360
Postoperative bile leakage (n = 18) 3.742 (1.310–10.685) .014 1.678 (0.319–8.828) .541
Relaparotomy because of short-term complications (n = 5) 4.590 (0.741–28.445) .102

Binary logistic regression where applicable.

Fig. 2. Stricture formation after hepaticojejunostomy in a tertiary center and subsequent treatment. HJ, hepaticojejunostomy; PTC, percutaneous transhepatic cholangiography.

Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
 ARTICLE IN PRESS
6 K.A.C. Booij et al. / Surgery ■■ (2018) ■■–■■
were successfully treated by subsequent dilatations. The remain-
ing 14 patients are discussed in the next paragraph.
Three patients did not need a percutaneous dilatation proce-
dure. One of these patients underwent percutaneous transhepatic
cholangiography in the referring hospital to diagnose stricture for-
mation, but during follow-up at our center, the patient became
asymptomatic. The second patient was diagnosed with a stricture
involving a segment 5 biliary duct but turned asymptomatic after
antibiotic treatment. The third patient was diagnosed with a stric-
ture in the referring hospital and was also successfully managed with
antibiotics. Two years later, during last follow-up, this patient was
still asymptomatic.
Recurrent strictures
In 14 of the 33 patients (42.4%) treated by dilatation, a recur-
rent stricture occurred after a median of 22.7 months (IQR 10.7–
52.2). These strictures were treated by direct surgical revision (n = 1)
or additional dilatation (n = 13). In 2 of the latter 13 patients, the
recurrent stenosis persisted after multiple dilatations and revi-
sion of the HJ was performed after 61 months and 28 months after
the initial HJ, respectively.
Overall, surgical revision of the HJ was performed in 4 patients
(1.4%) after a median of 17.7 months (IQR 6.8–52.7 months) without
additional mortality.
In 5 of the 33 patients (15.2%), a third episode of stricture for-
mation occurred after a stricture-free interval of a median of 57.6
months (IQR 27.6–96.4 months). These patients were treated with
a third period of percutaneous dilatations (n = 4). In 1 patient, the
biliary segment 6 duct was ultimately sclerosed with alcohol to
prevent recurrent cholangitis.
Discussion
The present series is one of the largest cohorts of HJ for BDI with
an extensive median long-term follow-up of 10 years. In this series,
stricture rate was higher in intermediate repair (14–90 days) versus
late repair (>90 days). Timing of surgery did not influence short-
and long-term outcomes after HJ.
Male gender was the only independent risk factor for HJ stric-
ture. Initial percutaneous management of HJ strictures was feasible
in 90% of patients and was the definitive treatment in almost 60%
of patients.
Throughout the literature a persistent debate exists about the
definition of early, intermediate, and late HJ reconstruction after BDI.
Definitions of reported timing range from fewer than 2–45 days for
early repair,8-10,20-22 to between 8 days–6 weeks for intermediate
repair 9 and more than 2 days to more than 6 weeks for late
repair.8-10,20,21 Because of the lack of an international definition an
adequate comparison of outcomes is not possible. In a previous study
from our center, we used 6 weeks to determine the early repair
period, which seems to be too long in view of increased risk of sepsis
and other BDI-related complications after 1–2 weeks.11 Therefore,
we adjusted the early repair period in this study to a period of up
to 2 weeks.12
Male gender was shown to be an independent risk factor for stric-
ture formation. To our knowledge, this has not been examined before.
Another recent study showed male gender to be a risk factor for
complications after HJ in univariable analysis.23 Survival after BDI
has previously been shown to be worse in men.5,24
In the present study, no significant difference was found in terms
of stricture rate for early (<14 days) versus intermediate and late
repair. This cannot be fully explained, but it may be caused by se-
lection bias. This result does not comply with our former results,
in which early repair was a negative predictor for outcome. However,
as discussed earlier, a different definition of timing was used.
Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
In another study, Kirks et al21 showed no effect of timing on
outcome, but a different definition on early and late repair was used.
Thomson et al25 showed a similar outcome between early repair (<2
weeks) and delayed repair (>6 weeks) for patients treated in an ex-
perienced center. Perera et al10 showed that immediate and early
repairs (<21 days) done by nonspecialist surgeons were indepen-
dent risk factors for recurrent cholangitis, restricturing, redo
reconstructions, and overall morbidity. Immediate and early repair
after BDI resulted in comparable, if not better, long-term out-
comes compared with late repair when performed by hepato-
pancreato-biliary surgeons. In contrast, Stilling et al8 found early HJ
(<2 weeks) to be a considerable risk factor for long-term compli-
cations and mortality compared with late HJ (>2 weeks). Iannelli
et al20 showed that timing of repair influences the rate of a neces-
sary, second procedure and concluded that a late repair is the
preferred option. In the largest cohort of 614 patients who under-
went HJ for BDI, analyzing primary versus secondary repair, the
authors found that sepsis control is a significant protective factor
for complications and anastomotic failure after primary repair.9 We
also believe that the clinical situation and in particular the pres-
ence of sepsis, biliary leakage/collections, perihepatic abscesses, and
concomitant vascular injury are more important than timing alone.
Timing in terms of days or weeks is probably a far less impor-
tant factor if these conditions are achieved and further delay of a
definitive treatment is avoided to improve clinical recovery and
quality of life. A recent study from Rystedt and Montgomery26
showed that immediate intraoperative repair demonstrated far better
results in terms of quality of life. But their interpretation is that their
unselected cohort of BDIs comprises a higher proportion of minor
injuries, detected intraoperatively, than in previous reports. Dageforde
et al27 demonstrated that early repair by a hepatobiliary surgeon
(<6 weeks) was associated with lower costs, earlier return to normal
activity, and better quality of life. But again, it might also be related
to the condition of the patient. Therefore, we should not only
compare early versus delayed repair but also consider different
patient circumstances in the timing for surgery.
In our study, we found no difference in the presence of preop-
erative sepsis before reconstruction. In the largest reconstructive
study after BDI (discussed earlier), it was shown that after second-
ary repair, intermediate timing of repair was a predictive risk factor
for complications, with no association with anastomotic failure. Their
timing of repair after injury was defined as early (<7 days), inter-
mediate (8 days–6 weeks) and late (>6 weeks). Anastomotic failure
was not defined as pure stricture formation of the HJ, but rather
as any episode of cholangitis (14%), biliary stricture (13%), and/or
jaundice after surgical repair, requiring endoscopic, radiological, or
surgical treatment. Stricture formation in the present study was com-
parable with this cohort and other publications.7,9
In our study, BDI was diagnosed after a median of 6 days after
LC in the referring hospitals. This is quite long. Intraoperative de-
tection using intraoperative cholangiography might have reduced
the incidence of major BDI,28-31 might have led to earlier diagnosis
of BDI and in the present of BDI early treatment or referral. This
might also leads to a better quality of life.26
Before referral, 24.6% of the patients had undergone a surgical
reintervention. This percentage is higher compared with our pre-
vious study and is reason for concern because reconstruction or
repair by nonspecialist surgeons is an important risk factor for worse
outcome.10,11
End-to-end anastomosis might be considered as primary treat-
ment for intraoperatively detected transection of the bile duct
without extensive tissue loss. Postoperative complications such as
a stricture can be adequately managed by endoscopic stenting in
the majority of patients (66%).32 In the remaining patients, recon-
structive surgery after complicated end-to-end anastomosis is a
procedure with excellent results.
 ARTICLE IN PRESS
K.A.C. Booij et al. / Surgery ■■ (2018) ■■–■■
Compared with other studies, our long-term surgical revision rate
(1.4%) was relatively low.8,9 Most HJ strictures (58%) were success-
fully treated by percutaneous dilatation. Literature on PTC treatment
for HJ strictures, especially after BDI, is scarce, with reported success
rates of 45–71%.8,17,33-35 Because percutaneous transhepatic dilata-
tion is nowadays easily accessible in expert centers with good
outcomes, an initial less-invasive treatment is preferred to
surgical revision. Our group strongly supports percutaneous
transhepatic dilatation as initial treatment for HJ stenosis. However,
subsequent dilatation is also suitable for recurrent stenosis and can
be considered as a step-up approach to surgical revision.
The present study has some strengths and limitations. The
strength is that it is one of the largest cohorts of BDI patients treated
by reconstruction in a single tertiary center with substantial long-
term follow-up. Even though part of this cohort was previously
reported,5,11 the present study prospectively included more recent
patients after HJ, a longer follow-up time, and new information on
stricture formation. A limitation of this retrospective study was that
detailed information on the extent of concomitant vasculobiliary
injury of patients, which might be an important factor for devel-
opment of long-term stenosis, unfortunately is lacking.
In conclusion, reconstructive surgery in BDI patients is associ-
ated with good short- and long-term outcomes. Postoperative
anastomotic strictures are associated with male gender, sepsis, repair
before referral, and postoperative complications. Percutaneous di-
lation is generally successful, and revision of the anastomosis should
only be considered if repeated dilatation fails. An international,
uniform definition of timing of reconstruction, as well as a defini-
tion and the management of postoperative strictures, would facilitate
comparison of outcomes among international series of patients. In
the discussion about timing of reconstruction, various patient cir-
cumstances, such as drainage and control of sepsis, should be
considered as well.
References
1. Flum DR, Cheadle A, Prela C, Dellinger EP, Chan L. Bile duct injury during
cholecystectomy and survival in medicare beneficiaries. JAMA 2003;290:2168-73.
2. Boerma D, Rauws EA, Keulemans YC, et al. Impaired quality of life 5 years after
bile duct injury during laparoscopic cholecystectomy: a prospective analysis. Ann
Surg 2001;234:750-7.
3. de Reuver PR, Sprangers MA, Rauws EA, et al. Impact of bile duct injury after
laparoscopic cholecystectomy on quality of life: a longitudinal study after
multidisciplinary treatment. Endoscopy 2008;40:637-43.
4. de Reuver PR, Rauws EA, Bruno MJ, et al. Survival in bile duct injury patients
after laparoscopic cholecystectomy: a multidisciplinary approach of
gastroenterologists, radiologists, and surgeons. Surgery 2007;142:1-9.
5. Booij KAC, de Reuver PR, van Dieren S, et al. Long-term impact of bile duct injury
on morbidity, mortality, quality of life, and work related limitations. Ann Surg
2017 Apr 19;doi:10.1097/SLA.0000000000002258. Epub ahead of print.
6. Strasberg SM, Helton WS. An analytical review of vasculobiliary injury in
laparoscopic and open cholecystectomy. HPB (Oxford) 2011;13:1-14.
7. AbdelRafee A, El-Shobari M, Askar W, Sultan AM, El Nakeeb A. Long-term
follow-up of 120 patients after hepaticojejunostomy for treatment of post-
cholecystectomy bile duct injuries: a retrospective cohort study. Int J Surg
2015;18:205-10.
8. Stilling NM, Fristrup C, Wettergren A, et al. Long-term outcome after early repair
of iatrogenic bile duct injury. A national Danish multicentre study. HPB (Oxford)
2015;17:394-400.
9. Dominguez-Rosado I, Sanford DE, Liu J, Hawkins WG, Mercado MA. Timing of
surgical repair after bile duct injury impacts postoperative complications but
not anastomotic patency. Ann Surg 2016;264:544-53.
Please cite this article in press as: Klaske A. Booij, et al., Long-term follow-up and risk factors for strictures after hepaticojejunostomy for bile duct injury: An analysis of surgical
and percutaneous treatment in a tertiary center, Surgery (2018), doi: 10.1016/j.surg.2018.01.003
7
10. Perera MTPR, Silva MA, Hegab B, et al. Specialist early and immediate repair of
post-laparoscopic cholecystectomy bile duct injuries is associated with an
improved long-term outcome. Ann Surg 2011;253:553-60.
11. de Reuver PR, Grossmann I, Busch OR, Obertop H, van Gulik TM, Gouma DJ.
Referral pattern and timing of repair are risk factors for complications after
reconstructive surgery for bile duct injury. Ann Surg 2007;245:763-70.
12. Gouma DJ, Callery MP. Bile duct injury: examining results of early repair by the
index surgeon. HPB (Oxford) 2017;19:1-2.
13. Strasberg SM, Hertl M, Soper NJ. An analysis of the problem of biliary injury
during laparoscopic cholecystectomy. J Am Coll Surg 1995;180:101-25.
14. Bergman JJ, van den Brink GR, Rauws EA, et al. Treatment of bile duct lesions
after laparoscopic cholecystectomy. Gut 1996;38:141-7.
15. Bismuth H, Majno PE. Biliary strictures: classification based on the principles
of surgical treatment. World J Surg 2001;25:1241-4.
16. de Reuver PR, Rauws EA, Vermeulen M, Dijkgraaf MGW, Gouma DJ, Bruno MJ.
Endoscopic treatment of post-surgical bile duct injuries: long term outcome and
predictors of success. Gut 2007;56:1599-605.
17. Janssen JJ, van Delden OM, van Lienden KP, et al. Percutaneous balloon dilatation
and long-term drainage as treatment of anastomotic and nonanastomotic benign
biliary strictures. Cardiovasc Intervent Radiol 2014;37:1559-67.
18. Blumgart LH. Hilar and intrahepatic biliary enteric anastomosis. Surg Clin North
Am 1994;74:845-63.
19. Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien-Dindo classification of
surgical complications: five-year experience. Ann Surg 2009;250:187-96.
20. Iannelli A, Paineau J, Hamy A, Schneck A, Schaaf C, Gugenheim J. Primary versus
delayed repair for bile duct injuries sustained during cholecystectomy: results
of a survey of the Association Francaise de Chirurgie. HPB (Oxford) 2013;15:611-
6.
21. Kirks RC, Barnes TE, Lorimer PD, et al. Comparing early and delayed repair of
common bile duct injury to identify clinical drivers of outcome and morbidity.
HPB (Oxford) 2016;18:718-25.
22. Sicklick JK, Camp MS, Lillemoe KD, et al. Surgical management of bile duct injuries
sustained during laparoscopic cholecystectomy: perioperative results in 200
patients. Ann Surg 2005;241:786-92, discussion 793-5.
23. Ismael HN, Cox S, Cooper A, Narula N, Aloia T. The morbidity and mortality of
hepaticojejunostomies for complex bile duct injuries: a multi-institutional
analysis of risk factors and outcomes using NSQIP. HPB (Oxford) 2017;19:
352-8.
24. Halbert C, Altieri MS, Yang J, et al. Long-term outcomes of patients with common
bile duct injury following laparoscopic cholecystectomy. Surg Endosc
2016;30:4294-9.
25. Thomson BNJ, Parks RW, Madhavan KK, Wigmore SJ, Garden OJ. Early specialist
repair of biliary injury. Br J Surg 2006;93:216-20.
26. Rystedt JML, Montgomery AK. Quality-of-life after bile duct injury: intraoperative
detection is crucial. A national case-control study. HPB (Oxford) 2016;18:1010-6.
27. Dageforde LA, Landman MP, Feurer ID, Poulose B, Pinson CW, Moore DE. A
cost-effectiveness analysis of early vs late reconstruction of iatrogenic bile duct
injuries. J Am Coll Surg 2012;214:919-27.
28. Ford JA, Soop M, Du J, Loveday BPT, Rodgers M. Systematic review of
intraoperative cholangiography in cholecystectomy. Br J Surg 2012;99:160-7.
29. Pesce A, Piccolo G, La Greca G, Puleo S. Utility of fluorescent cholangiography
during laparoscopic cholecystectomy: a systematic review. World J Gastroenterol
2015;21:7877-83.
30. Buddingh KT, Nieuwenhuijs VB, van Buuren L, Hulscher JBF, de Jong JS, van Dam
GM. Intraoperative assessment of biliary anatomy for prevention of bile duct
injury: a review of current and future patient safety interventions. Surg Endosc
2011;25:2449-61.
31. Buddingh KT, Weersma RK, Savenije RAJ, van Dam GM, Nieuwenhuijs VB. Lower
rate of major bile duct injury and increased intraoperative management of
common bile duct stones after implementation of routine intraoperative
cholangiography. J Am Coll Surg 2011;213:267-74.
32. de Reuver PR, Busch ORC, Rauws EA, Lameris JS, van Gulik TM, Gouma DJ.
Long-term results of a primary end-to-end anastomosis in peroperative detected
bile duct injury. J Gastrointest Surg 2007;11:296-302.
33. Misra S, Melton GB, Geschwind JF, Venbrux AC, Cameron JL, Lillemoe KD.
Percutaneous management of bile duct strictures and injuries associated with
laparoscopic cholecystectomy: a decade of experience. J Am Coll Surg
2004;198:218-26.
34. Cantwell CP, Pena CS, Gervais DA, Hahn PF, Dawson SL, Mueller PR. Thirty years’
experience with balloon dilation of benign postoperative biliary strictures:
long-term outcomes. Radiology 2008;249:1050-7.
35. Ramos-De la Medina A, Misra S, Leroy AJ, Sarr MG. Management of benign biliary
strictures by percutaneous interventional radiologic techniques (PIRT). HPB
(Oxford) 2008;10:428-32.