Journal of Pediatric Surgery (2011) 46, 209–213

www.elsevier.com/locate/jpedsurg

Hepaticoduodenostomy vs hepaticojejunostomy for

reconstruction after resection of choledochal cyst
Matthew T. Santore, Brittany J. Behar, Thane A. Blinman, Edward J. Doolin,
Holly L. Hedrick, Peter Mattei, Michael L. Nance, N. Scott Adzick, Alan W. Flake ⁎
The Department of Surgery, Children's Hospital of Philadelphia, Philadelphia, PA 19104, USA

Received 27 September 2010; accepted 30 September 2010

Key words:
Abstract
Choledochal cysts;
Purpose: Roux-en-Y hepaticojejunostomy (HJ) is currently the favored reconstructive procedure after
Hepaticojejunostomy;
resection of choledochal cysts. Hepaticoduodenostomy (HD) has been argued to be more
Hepaticoduodenostomy
physiologically and technically easier but is feared to have associated complications. Here we compare
outcomes of the 2 procedures.
Methods: A retrospective chart review identified 59 patients who underwent choledochal cyst resection
within our institution from 1999 to 2009. Demographic and outcome data were compared using t tests,
Mann-Whitney U tests, and Pearson χ2 tests.
Results: Fifty-nine patients underwent repair of choledochal cyst. Biliary continuity was restored by HD
in 39 (66%) and by HJ in 20 (34%). Open HD patients required less total operative time than HJ patients
(3.9 vs 5.1 hours, P = .013), tolerated a diet faster (4.8 days compared with 6.1 days, P = .08), and had a
shorter hospital stay (7.05 days for HD vs 9.05 days for HJ, P = .12). Complications were more
common in HJ (HD = 7.6%, HJ = 20%, P = .21). Three patients required reoperation after HJ, but only
one patient required reoperation after HD for a stricture (HD = 2.5%, HJ = 20%, P = .037).
Conclusions: In this series, HD required less operative time, allowed faster recovery of bowel function,
and produced fewer complications requiring reoperation.
© 2011 Elsevier Inc. All rights reserved.

The term choledochal cyst (CDC) describes cystic
dilation of various parts of the biliary tree that have in
common the potential for obstructive complications of the
common bile duct, pancreatic duct, and ultimately malignant
degeneration. In infancy, painless jaundice is the most
common presentation, whereas older children may present
with intermittent abdominal pain related to obstruction from
stones or sludge, with or without associated jaundice,
⁎ Corresponding author. Department of Surgery, Abramson Research
Center, Philadelphia, PA 19104-4318, USA. Tel.: +1 215 590 3671; fax: +1
215 590 3324.
E-mail address: flake@email.chop.edu (A.W. Flake).
pancreatitis, or cholangitis [1-4]. Choledochal cyst can be
diagnosed prenatally by ultrasound or magnetic resonance
imaging allowing early resection [5-9].
The treatment of CDCs has evolved from drainage
procedures to the current standard of complete excision of
the cyst with biliary reconstruction [10-13]. A variety of
creative surgical reconstructive procedures have been
described, but the 2 most common are reconstruction by
Roux-en-Y hepaticojejunostomy (HJ) or by hepaticoduode-
nostomy (HD). The HJ reconstruction seems favored by
most surgeons; however, there is minimal support in the
literature for this preference. The concerns related to HD
include the potential for cholangitis and bile gastritis, which

0022-3468/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jpedsurg.2010.09.092
210 M.T. Santore et al.

are theoretically addressed by HJ. However, reconstruction
by HD is technically easier, particularly in the context of
laparoscopic CDC resection, may be more physiologic,
avoids complications associated with the Roux-en-Y, and
allows postoperative endoscopic access to the anastomosis if
strictures or stones occur [14]. In view of these significant
potential advantages, and the minimal data supporting HJ
over HD in the literature, we decided to compare the results
of HD and HJ in our series of CDC resections.
1. Methods
A retrospective study was performed on all patients who
underwent excision of a CDC from August 1999 to July 2009
at the Children's Hospital of Philadelphia after obtaining
institutional review board approval. The data were collected
through retrospective review of the inpatient and outpatient
electronic medical record by 2 independent reviewers and
compared for consistency. The following data were collected:
presenting symptoms, complications of the disease, diagnos-
tic modality, size and type of the cyst, operative time, blood
loss, operative complications, duration of procedure, return to
regular diet, and postoperative complications. For descriptive
analysis, median and standard deviation were calculated for
each variable. For summary statistics, independent t tests,
Mann-Whitney U tests, and Pearson χ2 tests were used to
compare data from HD and HJ groups. The level of
significance (α level) was set at .05. A χ2 test was used for
any binary data with the α level set at .05.
2. Results
Of 59 total patients undergoing resection of CDC, Roux-
en-Y HJ was performed on 20 (34%) patients and the
remaining 39 (66%) underwent HD. The choice of
performing an HJ or HD was by surgeon preference. Most
of the operative procedures were performed by a single
surgeon (AWF) who performed both operative techniques.
The choice of operative technique evolved toward HD over
the duration of the series, and it was largely based on the ease
of approximation of the duodenum to the hepatic confluence
after performance of an extensive Kocher maneuver (ie, when
the anatomy allowed a tension-free HD anastomosis, an HD
was performed). There were no significant differences in the
male-to-female ratio, the median age at diagnosis, or the
median age at surgery in the HJ vs HD groups, respectively
(Table 1). Similar percentages of HJ and HD patients were
prenatally diagnosed (25% vs 21%). Patients who were
symptomatic before diagnosis presented with pain, nausea/
vomiting, fever, jaundice, acholic stools, pancreatitis, and/or
diarrhea. Hepaticojejunostomy patients had higher presenting
total bilirubin levels and had a higher incidence of

Table 1 Demographic data for HJ and HD patients

HJ Roux-En-Y HD Total
Ethnicity
White 15 (75%) 23 (59%) 38 (64%)
African 4 (20%) 9 (23%) 13 (22%)
American
Asian 0 (0%) 4 (10%) 4 (7%)
Hispanic 1 (5%) 1 (3%) 2 (3%)
Other 0 (0%) 2 (5%) 2 (3%)
Sex
Male 4 (20%) 13 (33%) 17 (29%)
Female 16 (80%) 26 (67%) 42 (71%)
Type of cyst
I 17 (85%) 34 (87%) 51 (86%)
II 1 (5%) 2 (5%) 3 (5%)
III 0 (0%) 1 (3%) 1 (1.6%)
IV 2 (10%) 2 (5%) 4 (6.7%)
V 0 (0%) 0 (0%) 0 (0%)
Median age at diagnosis 3 (IQR 0.3-4.0) 5.5 (IQR 1.1-11.0) P = .243
Median age at surgery 4 (IQR 1.1-4.0) 5.5 (IQR 1.1-11.0) P = .129
Diagnosed in utero 5 (25%) 8 (21%) P = .563
Symptomatic before diagnosis 14 (70%) 26 (67%) P = .795
Preoperative cholangitis 3 (15%) 1 (3%) P = .072
Average preoperative total 3.71 ± 5.46 1.75 ± 2.45 P = .618
bilirubin (mg/dL)
Previous operations 4 (20%) 7 (18%) P = .848
IQR indicates interquartile range.
HD vs HJ reconstruction after resection of choledochal cyst 211

preoperative cholangitis than HD patients, but these differ- Table 3 Postoperative findings
ences were not statistically significant (Table 1). The follow- HJ HD Significance
up period ranged from 3 months to 6 years (mean, 2.3 years) Roux-En-Y
for the HD patients and 4 months to 9 years (mean, 3.5 years)
Average length 9.05 ± 5.8 7.05 ± 6.0 P = .12
for the HJ patients.
of stay (d)
Type I CDC, a cystic dilation involving the hepatic duct
Average postoperative 2.5 ± 0.9 2.59 ± 1.2 P = .745
and common bile duct, was the most common type of cyst in epidural duration (d)
both HD and HJ groups (85% vs 87%). The size of the cysts Average postoperative 3.5 ± 2.5 3.5 ± 1.9 P = .923
in each group was similar with an average size of 3.2 and 2.2 NG tube duration (d)
cm for the HD and HJ groups, respectively (P = .154). The Average time to 6.1 ± 3.6 4.8 ± 5.6 P = .08
HJ patients had a significantly longer operative time when regular diet (d)
compared to the entire HD group (5.12 vs 4.15 hours, P = Postoperative leak 2 (10%) 2 (5%) P = .31
.038). However, the HD group includes 6 laparoscopic Postoperative stricture 1 (5%) 1 (2.6%) P = .45
procedures that required significantly longer times to Postoperative 1 (5%) 0 (0%) P = .33
complete than open HD (Table 2). If the laparoscopic HD cholangitis
Reoperation 4 (20%) 1 (2.5%) P = .037
procedures are excluded from the analysis, there is a larger
significant difference in operative time between the 2 open NG, Nasogastric.
procedures. Operative blood loss was not significantly
different between HJ and HD groups.
Postoperatively, there was a trend toward earlier dis- The complications within the HD group included a
charge in the HD patients excluding patients who developed superficial wound infection that resolved with antibiotics, a
complications. There was no difference between the 2 groups biliary leak diagnosed by Hepatobiliary iminodiacetic acid
in duration of epidural pain control. Importantly, the HD (HIDA) scan that was successfully managed nonoperatively,
group had a trend toward shorter time to resumption of a and a stricture and leak within one patient that required
regular diet than the HJ patients (HD = 4.8 days vs HJ = 6.1 reoperation. This patient developed biliary peritonitis
days; P = .08), but this was not significant and both groups requiring an exploratory laparotomy on postoperative day
had similar intervals of nasogastric decompression. When 15 after nonoperative management failed. During the
comparing postoperative complications, there were no reoperation, the anastomosis was intact, but the hepatic
significant differences in the rates of postoperative leak, biliary duct showed a small leak treated with fibrin seal and
infection, reoperations, and cholangitis between the 2 abdominal drainage. This patient subsequently did well after
groups. There were 3 (7.7%) of 39 compared to 4 (20%) of an Endoscopic Retrograde Cholangiopancreatography
20 patients who developed complications in the HD vs HJ (ERCP) dilatation of a minimal stricture of the distal hepatic
groups, respectively. However, the HJ group had a bile duct. The HJ group required 4 reoperative procedures
significantly greater rate of complications requiring reopera- because of complications of the initial surgery. Two of the
tions compared to the HD group (HJ 20% vs HD 2.5%, P = reoperations were not related to the biliary anastomosis. One
.037). Rates of postoperative stricture in the 2 groups were of these patients required a repair of an incisional hernia
not significantly different. There was only one patient within along the subcostal incision and the other patient, who had a
the entire study who developed cholangitis and that patient history of gastroschisis repair and a Ladd procedure,
underwent HJ (Table 3). developed an early postoperative bowel obstruction and
required an exploratory laparotomy with lysis of adhesions
and detorsion of a segmental volvulus. Of the 2 patients
requiring revision of the biliary anastomosis, one developed
Table 2 Perioperative characteristics
cholangitis from a high-grade obstructive stricture and
HJ HD Significance ultimately required 2 operative revisions of the HJ because
Roux-En-Y of several episodes of cholangitis. The other patient
Median size of 2.2 cm 3.2 cm P = .154 developed a biliary leak with peritonitis on the fifth
cyst (cm) (range, (range, postoperative day and at exploration was found to have a
0.8-0.6 cm) 0.4-12 cm) leak from a point on the left hepatic duct 1 cm proximal to the
Average operative 307 249 P = .038 HJ anastomosis. This was patched with adventitial tissue and
time total HD omentum and drained with resolution of the leak.
cases vs HJ (min)
Average operative 307 235 P = .013
time open HD vs 3. Discussion
HJ (min)
Average blood 31.8 ± 14.7 25.7 ± 12.7 P = .080 The debate surrounding biliary reconstruction by HJ vs
loss (mL) HD centers on the opinion that the HD reconstruction is more
frequently complicated by bile gastritis, cholangitis, and is
212
associated with a higher ongoing risk of cholangiocarcinoma
[15-17]. A review of the literature, however, finds surpris-
ingly little data supporting these concerns and there have
been relatively few studies published comparing HD to HJ.
In an early report, Todani et al compared 19 patients
undergoing HD and 11 undergoing HJ and found no
significant difference in biliary complications between the
2 procedures and advocated the HD procedure because of its
“more physiologic state” and fewer postoperative intestinal
complications [18]. Perhaps the most influential publication
is by Shimotakahara et al [19], which compared 28 patients
reconstructed by HJ with 12 patients reconstructed by HD
after CDC resection. They observed complications of bile
gastritis in 4 of 12 HD patients compared with 0 of 28 HJ
patients and had 2 postoperative adhesion-related bowel
obstructions in the HJ group with none in the HD group.
They concluded that HJ was the surgical reconstruction of
choice because of the high incidence of duodenogastric
reflux after HD [19]. Supporting their contention is a study
by Takada et al [20] in which 3 patients who underwent HD
reconstruction and 5 who underwent HJ reconstruction were
compared by monitoring with a Bilitec probe (Medtronic,
Minneapolis, MN) and were endoscoped to assess the
presence and severity of duodenogastric reflux. Although
none of the patients were symptomatic, all 3 of the HD
patients and none of the HJ patients had chemical and
endoscopic evidence of duodenogastric reflux. However,
biopsies showed only superficial gastritis in both groups.
Finally, a report by Todani et al [21] of hilar cholangio-
carcinoma 19 years after CDC excision and HD reconstruc-
tion has been influential in favoring HJ over HD.
Our results differ from those of Shimotakahara et al as we
observed no incidence of symptomatic bile gastritis. The
reasons for this difference are unclear as is the reason that
patients undergoing an HD reconstruction should have
greater amounts of bile gastritis than healthy patients. Our
construction of the HD anastomosis is at the junction of the
first and second portions of the duodenum performed after an
extensive Kocher maneuver to prevent any tension on the
anastomosis. This is well distal to the pylorus and should not
impact pyloric function or gastric emptying significantly.
Although we have not routinely endoscoped patients to
monitor for bile gastritis, it has not been a clinical issue in
this series. Similarly, we have not seen cholangitis in the
absence of stricture formation with either procedure. This is
in agreement with all large published series of CDC excision.
In circumstances where there is free bile drainage, the
incidence of cholangitis should be minimal with either
reconstruction. Finally, the question of whether HD entails
an increased ongoing risk of cholangiocarcinoma at the
hepatic duct confluence or of the intrahepatic ducts is more
difficult to address without very long-term follow-up.
However, most reports of intrahepatic carcinoma after
complete excision of CDC are associated with intrahepatic
ductal dilation, stones, strictures, and/or recurrent cholangitis
despite a Roux-en-Y HJ reconstruction in several cases [22].
M.T. Santore et al.
Of particular relevance to this discussion is the report of a
patient by Goto et al who underwent resection of a type Ia
cyst (no intrahepatic dilation) with development 10 years
later of an intrahepatic cholangiocarcinoma. She had a Roux-
en-Y HJ reconstruction [23]. Thus, there may be a field risk
of carcinoma of intrahepatic ducts in patients with
anomalous pancreaticobiliary junctions irrespective of the
type of reconstructive procedure that they undergo.
In this series we found few significant differences
between our outcomes with HD compared to HJ. It is clear
that HD reconstruction requires less operative time and is a
simpler procedure to perform than HJ. This is particularly
true of the laparoscopic procedure where the construction of
the Roux-en-Y can add significant difficulty to the procedure
and is often performed extracorporeally. There was also a
trend toward shorter times to enteral feeds and length of stay
after HD. Although these findings were not statistically
significant, they are likely to be real given the reduced bowel
manipulation required in the HD procedure. Importantly,
there did not appear to be an increased rate of bile leak,
stricture formation, or cholangitis in the HD group.
However, one of the bile leak patients who underwent an
HD illustrates one of the clear advantages of HD over HJ,
which is endoscopic accessibility of the anastomosis. If an
HD reconstruction is performed, the ability to endoscopically
place stents or dilate the HD can avoid the need for
reoperation when a leak or stricture occurs. Finally, our HJ
patients had a higher rate of reoperation than the HD group.
One of these was related to intestinal obstruction as has been
observed in other comparative series. The complications
related to Roux-en-Y conduits are well documented in the
literature and, aside from mechanical or adhesion-related
complications, include reports of fat malabsorption and
duodenal ulcers supporting the concept that HD drainage is
more physiologic and suggesting that a Roux-en-Y should be
avoided if there is an equally effective alternative [16,24].
Our results support HD as the preferred procedure for
biliary reconstruction after resection of CDC. In our view, the
advantages of relative simplicity, particularly with the
laparoscopic approach, avoidance of complications related to
the Roux-en-Y conduit, postoperative endoscopic accessibil-
ity of the anastomosis, and restoration of relatively physiologic
bile drainage outweigh the potential disadvantages of
this approach, which for the most part remain unproven.
References
[1] Miyano T, Yamataka A, Kato Y, et al. Hepaticoenterostomy after
excision of choledochal cyst in children: a 30-year experience with 180
cases. J Pediatr Surg 1996;31:1417-21.
[2] Miyano T, Yamataka A, Li L. Congenital biliary dilatation. Semin
Pediatr Surg 2000;9:187-95.
[3] Davenport M, Betalli P, D'Antiga L, et al. The spectrum of surgical
jaundice in infancy. J Pediatr Surg 2003;38:1471-9.
[4] de Vries JS, de Vries S, Aronson DC, et al. Choledochal cysts: age of
presentation, symptoms, and late complications related to Todani's
classification. J Pediatr Surg 2002;37:1568-73.
HD vs HJ reconstruction after resection of choledochal cyst
[5] Liu SL, Li L, Hou WY, et al. Laparoscopic excision of choledochal
cyst and Roux-en-Y hepaticojejunostomy in symptomatic neonates.
J Pediatr Surg 2009;44:508-11.
[6] Mishra A, Pant N, Chadha R, et al. Choledochal cysts in infancy and
childhood. Indian J Pediatr 2007;74:937-43.
[7] Lin WJ, Yu CY, Diau GY, et al. Use of magnetic resonance
cholangiopancreatography to diagnose neonatal congenital choledo-
chal cyst. J Formos Med Assoc 2005;104:441-3.
[8] Mackenzie TC, Howell LJ, Flake AW, et al. The management of
prenatally diagnosed choledochal cysts. J Pediatr Surg 2001;36:
1241-3.
[9] Todani T, Urushihara N, Morotomi Y, et al. Characteristics of
choledochal cysts in neonates and early infants. Eur J Pediatr Surg
1995;5:143-5.
[10] Ohi R, Yaoita S, Kamiyama T, et al. Surgical treatment of congenital
dilatation of the bile duct with special reference to late complications
after total excisional operation. J Pediatr Surg 1990;25:613-7.
[11] Singham J, Schaeffer D, Yoshida E, et al. Choledochal cysts: analysis
of disease pattern and optimal treatment in adult and paediatric
patients. HPB (Oxford) 2007;9:383-7.
[12] Cosentino CM, Luck SR, Raffensperger JG, et al. Choledochal duct
cyst: resection with physiologic reconstruction. Surgery 1992;112:
740-7 [discussion 747-8].
[13] Edil BH, Olino K, Cameron JL. The current management of
choledochal cysts. Adv Surg 2009;43:221-32.
[14] Liem NT, Dung LA, Son TN. Laparoscopic complete cyst excision
and hepaticoduodenostomy for choledochal cyst: early results in 74
cases. J Laparoendosc Adv Surg Tech 2009;19:87-90.
[15] Li MJ, Feng JX, Jin QF. Early complications after excision with
hepaticoenterostomy for infants and children with choledochal cysts.
Hepatobiliary Pancreat Dis Int 2002;1:281-4.
[16] Saing H, Han H, Chan KL, et al. Early and late results of excision of
choledochal cysts. J Pediatr Surg 1997;32:1563-6.
[17] Oweida SW, Ricketts RR. Hepatico-jejuno-duodenostomy reconstruc-
tion following excision of choledochal cysts in children. Am Surg
1989;55:2-6.
[18] Todani T, Watanabe Y, Mizuguchi T, et al. Hepaticoduodenostomy at
the hepatic hilum after excision of choledochal cyst. Am J Surg
1981;142:584-7.
[19] Shimotakahara A, Yamataka A, Yanai T, et al. Roux-en-Y
hepaticojejunostomy or hepaticoduodenostomy for biliary reconstruc-
tion during the surgical treatment of choledochal cyst: which is better?
Pediatr Surg Int 2005;21:5-7.
[20] Takada K, Hamada Y, Watanabe K, et al. Duodenogastric reflux
following biliary reconstruction after excision of choledochal cyst.
Pediatr Surg Int 2005;21:1-4.
[21] Todani TWY, Toki A, Hara H. Hilar duct carcinoma developed after
cyst excision followed by hepaticoduodenostomy. In: Koyanagi Y,
Aoki T, editors. Pancreaticobiliary maljunction. Tokyo: Igaku tosho
shuppan; 2002. p. 17-21.
[22] Todani T, Watanabe Y, Urushihara N, et al. Biliary complications after
excisional procedure for choledochal cyst. J Pediatr Surg 1995;30:
478-81.
[23] Goto N, Yasuda I, Uematsu T, et al. Intrahepatic cholangiocarcinoma
arising 10 years after the excision of congenital extrahepatic biliary
dilation. J Gastroenterol 2001;36:856-62.
213
[24] Yamataka A, Kobayashi H, Shimotakahara A, et al. Recommenda-
tions for preventing complications related to Roux-en-Y hepatico-
jejunostomy performed during excision of choledochal cyst in
children. J Pediatr Surg 2003;38:1830-2.
Discussion
John Raffensperger, MD (Sanibel, Fla)
Dr Raffensperger, MD: Stricture, leak, and cholangitis. It
seems to me that is a lot of complications. Did you do
frozen sections at the proximal end of the resection to
see that you had normal mucosa? Sometimes you have
to go back all the way to the junction of the right and
left hepatic ducts. And in other areas of the body such
as the ureterovesical junction and the esophagogastric
junction it does not pay to ignore the body's natural
sphincters, the natural valves that prevent reflux. First
of all, the draining bile way down in the jejunum is not
physiologic. Draining bile into the duodenum without a
valve is not physiologic. It seems to me that there is a
third alternative and that is to take an isolated segment
of jejunum with a little intussuscepted valve and then
you protect your system from cholangitis if you make
sure you have normal mucosa at the proximal end you
won't have strictural leak.
Your follow-up is very brief. Three years isn't very much
when these patients have a life expectancy of 70, maybe
even 80 years, like me.
Dr Santore (response): As far as some of our complications,
as far as biliary gastritis, we do an extensive Kocher
maneuver of the duodenum and do an anastomosis fairly
far down on the duodenum. I appreciate your comments
as far as alternatives for different types of operations.
Unidentified Speaker: I have one quick question. In your
abstract you eliminated the laparoscopic hepaticoduode-
nostomies from your op time and talked about whether or
not it was shorter. Do you do the same thing for your
feeding time? Did you break out the …
Dr Santore: I did.
Speaker: So that feeding time is actually all the opens?0
Dr Santore: Yes, correct.