1. ORIGINAL ARTICLES
Biliary Reconstruction Using a Side-to-Side
Choledochocholedochostomy With or Without T-Tube in
Deceased Donor Liver Transplantation
A Prospective Randomized Trial
Sascha Weiss, MD, Sven-Ch Schmidt, MD, Frank Ulrich, MD, Andreas Pascher, MD, PhD,
Guido Schumacher, MD, PhD, Martin Stockmann, MD, PhD, Gero Puhl, MD, PhD,
Olaf Guckelberger, MD, PhD, Ulf P. Neumann, MD, PhD, Johann Pratschke, MD, PhD,
and Peter Neuhaus, MD, PhD
reconstruction is still the most challenging surgical step. Biliary
Objective: The biliary anastomosis is still one of the major causes for
complications continue to be a major cause of morbidity in liver
morbidity after orthotopic liver transplantation. The optimal method of
transplant recipients with an incidence up to 30% and a mortality
reconstruction remains controversial. The aim of the study was to assess
rate up to 10%. The most common biliary complications are bile
biliary complications after liver transplantation using a choledochochole-
leaks in the early postoperative period and biliary strictures typically
dochostomy with or without a temporary T-tube.
developing over long-term.1– 6
Background Data: Several reports have suggested that biliary reconstruc-
However extended donor criteria and the increasing use of
tion without T-tube is a safer method with a lower rate of biliary complica-
suboptimal organs have highlighted the relevance of donor-associ-
tions compared with T-tube insertion.
ated risk factors. There is growing evidence that these risk factors
Methods: A total of 194 recipients of deceased donor liver grafts were
have a significant influence on graft function and complication rates
randomized. In group 1 the biliary reconstruction was performed by side-to-
of the biliary anastomoses. This is supported by the fact that an
side choledochocholedochostomy with (n 99) and in group 2 (n 95)
increasing rate of retransplantations is necessary due to biliary
without a T-tube. The T-tube was removed after 6 weeks.
associated complications.7
Results: The overall biliary complication rate was significantly increased in
The risk of biliary complications is related to the technique of
group 2 (P 0.0005). Biliary leaks occurred in 5 patients in group 1 and in
bile duct reconstruction, the type of liver transplant performed, and
9 patients in group 2 (5.05% vs. 9.47%; P 0.2756 ns). Anastomotic
the usage of an external or internal drainage of the bile duct
strictures of the bile duct were seen in 7 patients in group 1 and in 8 patients
anastomoses. Reduced size graft, split liver transplantation, and
in group 2 (7.07% vs. 8.42%; P 0.7923 ns). Two of the patients in group
living donor liver transplantation are associated with an increased
1 and 5 patients in group 2 developed an ischemic type biliary lesion (2.02%
rate of biliary complications with a reported incidence up to 60%.
vs. 5.26%; P 0.2716 ns). The rate of reoperations was comparable in both
After full size liver transplantation the anastomoses of the bile duct
groups. The rate of invasive interventions was higher in the group without
may be routine, nevertheless the limited borderline arterial perfusion
T-tubes (9% vs. 18%, P ns), as was the rate of cholangitis (5% vs. 11%.
of the bile duct may cause significant complications. This seems to
P ns) and pancreatitis (4% vs. 14%, P 0.0218). No complications after
be of increasing relevance regarding the fact of an increasing
removal of the T-tube were observed.
acceptance of old and marginal donor organs.8
Conclusion: This study is a large prospective randomized trial to assess
The insertion of a T-tube has been the subject of controversy.
biliary complications that occur following liver transplantation, after anato-
So far a few prospective randomized trials examined the impact of
mizing the bile duct with or without T-tubes. A significant increased rate of
T-tube use after liver transplantation. These trials revealed mixed
complications in the group without T-tube insertion was observed. In
results, concluding that the usage of T-tubes is associated with a
summary, our results indicate that the usage of T-tubes is safe and an
higher rate of complications after liver transplantation. Still a sub-
excellent tool for the quality control of biliary anastomoses.
stantial number of centers prefer the routine use of a T-tube for the
(Ann Surg 2009;250: 766 –771) simple diagnostic access to the bile duct systems, more recently a
growing number of center have been advocating biliary anastomoses
without stenting or the use of a T-tube.9,10
To evaluate the effect of T-tube insertion after deceased
donor liver transplantation and its implications on the morbidity,
A dvances in immunosuppressive protocols, organ preservation,
and perioperative management have significantly improved pa-
tient and graft survival after liver transplantation. Biliary tract
mortality, and long-term effects we designed a prospective random-
ized trial.
MATERIALS AND METHODS
From the Department of General, Visceral and Transplantation Surgery, Charite,
´ Patients
Campus Virchow Clinical Center, Universitatsmedizin Berlin, Germany.
¨
S.W. and S.C.S. contributed equally to this report. From 2005 to 2007 in this single center study, 194 patients
Reprints: Peter Neuhaus, MD, PhD, Department of General, Visceral and Trans- were prospectively randomized in 2 groups. The first group of
plantation Surgery, Charite, Campus Virchow Clinical Center, Augusten-
´ patients were assigned to a side-to-side choledochocholedochos-
burger Platz 1, 13353 Berlin, Germany. E-mail: peter.neuhaus@charite.de.
Copyright © 2009 by Lippincott Williams & Wilkins
tomy with T-tube (n 99), the second using the same surgical
ISSN: 0003-4932/09/25005-0766 technique without (n 95). The T-tube was removed after 6 weeks.
DOI: 10.1097/SLA.0b013e3181bd920a Inclusion criteria were patients receiving a deceased full size liver
766 | www.annalsofsurgery.com Annals of Surgery • Volume 250, Number 5, November 2009

2. Annals of Surgery • Volume 250, Number 5, November 2009 Biliary Complications After Liver Transplantation
graft and recipient age older than 18 years. Exclusion criteria were tested with saline after fixing the T-tube with 2 stitches. The T-tube
patients under 18 years of age, retransplantation, the need of hepati- was perforated through the abdominal wall and connected with a
cojejunostomies, split or reduced size grafts, living donation, portal collecting bag.
thrombosis, and the need of an arterial interposition graft. In this
trial the surgical procedure was highly standardized, the transplan- Handling T-Tube
tations were performed exclusively by 4 surgeons with an individual X-ray cholangiography was performed 5 days after transplan-
experience of over 100 liver transplantations at the time of study tation. If the postoperative course was uneventful, the cholangiog-
initiation. Randomization was performed with sealed envelopes raphy demonstrated a sufficient outflow into the duodenum and the
before transplantation and the result were told the surgeon at the bilirubin levels decreased adequately, the T-tube was closed subse-
time he started suturing the biliary anastomoses. All types of quently. After 6 weeks, a second x-ray cholangiography was per-
complications were assessed including minor events, laboratory formed routinely, and the T-tube was removed, if no abnormalities
values were assessed within the first 7 days after transplantation were detected.
daily, thereafter twice a week, afterward on a outpatient basis after
3, 6, and 12 months. Additionally, the need of diagnostic procedures Definition of Complications
as well as the rate of diagnostic interventions was documented. All Bile leakages were defined as presence of a significant intra-
patients received ursodeoxycholic acid (250 mg) twice a day for 6 abdominal collection of bilirubin requiring ultrasound or radiologic
weeks after transplantation to increase bile secretion and reduce bile guided puncture. Alternatively, the leakage was proven by endo-
viscosity. The transplantation unit of the department provided fol- scopic retrograde cholangiography (ERC) or x-ray cholangiography.
low-up of the patients. This study was approved by the ethics Biliary stenosis were confirmed in all cases by x-ray cholan-
committee of the Charite, Universitatsmedizin Berlin, Germany.
´ ¨ giography or ERC.
Donor Surgical Technique Ischemic type biliary strictures were diagnosed by ERC and
The donor operation and harvesting procedure were per- were characterized by intrahepatic strictures in the absence of a
formed using standard techniques, 80% of the transplanted livers hepatic artery thrombosis.
were shipped, and local teams from the department of surgery The diagnosis of a papillary stenosis was based on the
retrieved 20%. University of Wisconsin solution was used for organ elevation of bilirubin levels and cholestatic enzymes and confirmed
preservation in all cases. by ERC in all cases.
Cholangitis was defined by elevated infectious parameters,
Recipient Surgical Technique the presence of fever in combination with elevated cholestatic
Total hepatectomy without caval preservation was performed parameters.
in the majority of recipients (Table 1). Venovenous bypass was not
used. Hepatic arteries and portal veins were anastomosed before Statistical Analysis
simultaneous reperfusion and in average 500 mL perfusate and Data are expressed as mean SD (standard deviation) in text
blood were discharged via the inferior vena cava. The grafts were or tables and presented as mean SEM (standard error of the mean)
not flushed with crystalloids before reperfusion. A side-to-side in figures. For single comparisons, normally distributed data were
anastomosis was performed with preserving the connective tissue. analyzed using unpaired, 2-tailed T-test, nonparametric Mann–
The end of the bile ducts were ligated (3-0 Vicryl). After finishing Whitney test, evaluated non-normally distributed data. Categorical
the anastomoses of the back wall with running suture (6-0 Prolene) variables were expressed as number (ratio/percent) and compared
in group 1 a small rubber T-tube (2.5 mm) was tailored and fit on a using the 2 test or Fisher exact test. P 0.05 (2-sided) were
special bile duct probe (Waldemar Link, Hamburg, Germany). The considered statistically significant.
probe was brought out through a small hole of the recipients’ A power calculation to detect a minimum 50% difference in
common bile duct. The anterior wall of the anastomoses is then the overall complication rate with a Type 1 error (a) of 0.05 and 80%
closed with a running suture commencing on both corners toward power, suggested that n 90 per group were required to demon-
the middle of the anastomoses. Patency of the anastomoses was strate significant differences.
TABLE 1. Recipient Characteristic and Intraoperative Data
T-Tube (n 99) Without T-Tube (n 95)
Mean SD ( SE) Mean SD ( SE) P Test
Recipient age (yr) 53.34 9.81 ( 0.99) 55.34 6.66 ( 0.68) 0.4226 Mann-Whitney
Sex (male/female) 69/30 60/35 0.3638 Fisher exact test
MELD score 16.97 8.78 ( 0.89) 16.62 8.43 ( 0.87) 0.9121 Mann-Whitney
Donor age (yr) 53.62 15.34 ( 1.61) 57.68 15.02 ( 1.57) 0.0618 Mann-Whitney
ICU (d) 5.57 5.08 ( 0.52) 5.43 6.36 ( 0.66) 0.6467 Mann-Whitney
CIT (min) 597 159 ( 6) 559 178 ( 18) 0.1142 Mann-Whitney
Operation time (min) 308 63 ( 6) 314 69 ( 7) 0.7168 Mann-Whitney
WIT (min) 43.2 6.6 ( 0.8) 45.0 6.6 ( 0.8) 0.3712 Mann-Whitney
RBC 5.3 4.3 ( 0.5) 6.0 5.0 ( 0.6) 0.4737 Mann-Whitney
FFP 17.0 7.4 ( 0.9) 18.0 9.4 ( 1.1) 0.7943 Mann-Whitney
Piggy back (yes/no) 72 (73%)/27 (27%) 74 (78%)/21 (22%) 0.4119 Fisher exact test
ICU indicates intensive care unit; CIT, cold ischemia time; WIT, warm ischemia time; RBC, red blood cell transfusion; FFP, fresh frozen plasma;
MELD, Model for end-stage liver disease.
© 2009 Lippincott Williams & Wilkins www.annalsofsurgery.com | 767

3. Weiss et al Annals of Surgery • Volume 250, Number 5, November 2009
RESULTS
Donor Characteristics
There were no statistical differences between both groups
regarding donor age, donor gender, hypotension, cardiac resuscita-
tion, and perioperative cardiac arrests as well as the fluids (including
infusions, albumin, etc) and catecholamines applied in both groups
(Table 1). The preservation time including cold and warm ischemia
time was equal in both donor groups (596 159 minutes in group
1 vs. 559 178 minutes in group 2). The time spent at the intensive
care unit was comparable in both groups (5.57 5.08 days in group
1 vs. 5.53 6.36 days in group 2). The operation time was
comparable in both groups (308 63 minutes in group 1 vs. 314
69 minutes in group 2).
Recipient Characteristics
No statistical differences were observed in patients receiving
either a graft from treated or untreated donors with regard to the
recipient age, donor gender, and diagnosis (Table 1). The recipient’s
status according to the Model for end-stage liver disease score did FIGURE 1. One-year patient survival is comparable between
not reveal significant differences between the groups (Model for both groups.
end-stage liver disease score average was 16.9 in group 1 vs. 16.6 in
group 2). The time for vascular anastomoses, defined as warm
ischemia time, was comparable in both groups (43 11 minutes in severe complications with pancreatitis as an consequence of neces-
group 1 vs. 45 9 minutes in group 2). sary ERC was higher in the group without T-tubes simply based on
the fact that these patients were more frequently in need of invasive
Immunosuppression diagnostic and therapeutic intervention compared with group 1
The immunosuppressive regimen in both groups was based (Table 2). Moreover the rate of leakages in group 2 lead to 4
on Tacrolimus or Cyclosporine A in addition with Mycophenolate reoperations with reanatomizing of the biliary anastomoses and
Mofetil and steroids. None of the patients received an induction placement of a T-tube to decompress the bile system. Reoperations
therapy. There were no significant differences in the immunosup- due to biliary complications were necessary in the group with
pressive regimen between both groups, over short and long term; T-tubes in only 1 case as the bile system was decompressed and 4 of
especially the dosage of steroids did not differ significantly. The rate these leakages occurred at the T-tube insertion. All these insuffi-
of hepatitis C virus reinfections after liver transplantation was ciencies could be treated conservatively with sufficient decompres-
comparable in both groups 3 months after transplantation. All sion of the bile system provided by the T-tube, leading to sponta-
patients received as prophylaxis against cholangitis 250 mg cipro- neous persistence of the insufficiency. We did not observe
floxacin twice daily. complications after removal of the T-tube after 6 weeks. A tempo-
Complication Rate rary increase of transaminases or cholestatic parameters within the
first 3 days after T-tube removal was not classified as complication.
The overall complication rate was significantly higher in the
group without T-tube (27% vs. 50%, Table 2). Moreover the rate of Long-Term Outcome
Patient and graft survival did not differ significantly up to 1
year (Fig. 1). A higher rate of ischemic type biliary lesions was
TABLE 2. Complication Rates observed at 1 year in group 2 without T-tube although not reaching
statistical significance (Table 2).
Without Without vs.
T-Tube T-Tube T-Tube P Cost Analysis
(n 99) (n 95) (Fisher Exact Test) The average in hospital costs between group 1 and 2 were
Leakage 5 (5.05%) 9 (9.47%) 0.2756 assessed and showed significant higher expenses in the group
Stenosis of 7 (7.07%) 8 (8.42%) 0.7923
without T-tube (P 0.43).
anastomosis Initial Graft Function and Rejection Rates
6 wk 4 (4.04%) 6 (6.32%) 0.5309 There were no differences in the initial graft function and
6 wk 3 (3.03%) 2 (2.11%) 1.0000 rejection rates within the follow-up period up to 1 year (Tables 1, 2).
ITBL at 1 yr 2 (2.02%) 5 (5.26%) 0.2716
Pancreatitis 4 (4.04%) 13 (13.68%) 0.0218 DISCUSSION
Reoperations 4 (4.04%) 4 (4.21%) 1.0000 The choice of biliary anastomoses is a major determinant for
Due to leakage/ 1 4 0.1728 the risk of biliary complications after orthotopic liver transplanta-
stenosis tion. To evaluate the benefit and risks of the usage of T-tubes we
Cholangitis 5 (5.05%) 11 (11.58%) 0.1206 designed a prospective randomized trial in 194 liver transplant
Complications/total 27 (27.27%) 50 (52.63%) 0.0005 recipients.
Rate of ERC 9 (9.09%) 18 (18.95%) 0.0615 However the insertion of a T-tube is discussed controversially
Rejection episodes 14 (14.14%) 18 (18.95%) 0.4403 in the surgical community, as its use can be associated with inherent
complications. Complication rates associated with T-tubes were
ITBL indicates ischemic type biliary lesions; ERC, endoscopic retrograde cholan- reported up to 50%. Most of these complications seem to be related
giography.
directly to the fact that a T-tube was inserted. These reports dis-
768 | www.annalsofsurgery.com © 2009 Lippincott Williams & Wilkins

4. Annals of Surgery • Volume 250, Number 5, November 2009 Biliary Complications After Liver Transplantation
couraged many centers of the further use of T-tubes and advocated However we could not confirm these high rates of cholangitis
biliary duct reconstruction without external bile drainage. However, which was 5% in the T-tube group and 11% without in our series.
the advantages of T-tubes are well known. The use of a T-tube One difference could be the definition of cholangitis, as in our trial
allows easy postoperative assessment of bile flow and bile quality a positive bacterial testing at the site of T-tubes without clinical and
for monitoring purposes. Additionally, it provides a safe radiologic laboratory findings of cholangitis did not qualify as cholangitis. It
access to the biliary anastomoses and anatomy. It is reported that the seems important which kind of complications is included in the
inherent decompression of the bile system associated with T-tubes assessment. We regard for example the imminent occurrence of
leads to reduced anastomotic insufficiencies as well as to a reduced pancreatitis as an important complication after ERC due to missing
incidence of anastomotic strictures. a T-tube. Pancreatitis after ERC has the potency of a disastrous
Three prospective randomized trials examine the advantages outcome and is therefore included as a complication in our trial.
and disadvantages of the usage of T-tubes are published. The first Another important point seems that reports showed that the
one reported higher rates of strictures in the group without T-tube, conversion rate from duct-to-duct anastomoses to hepaticojejunos-
as well as an increased rate of surgical revision of the biliary tomies, implicating major surgery in the non-T-tube group was
anastomoses performing subsequentially a hepaticojejunostomy. significantly higher than in the T-tube group. An indwelling T-tube
The number of patients included in this study was low and therefore stent was therefore considered to be useful for both, achieving the
the statistical power limited. Nevertheless these results are partially lowest possible rate of severe anastomotic stricture and to prevent
confirmed by our study also observing a low incidence of early any subsequent intervention.12 We did not perform any conversions
anastomotic strictures in the T-tube group. However, this finding of the anastomoses but had 4 revisions and reoperations in the group
was not significantly different for the early phase compared with the without T-tubes with subsequently insertion of a T-tube for decom-
group without drainage. pressing the bile duct system with an increased pressure in the bile
The second study randomized 90 patients in each group in a ducts as assumed reason for the anastomotic leakages.
multicenter trial and revealed a significant increase of biliary com- The majority of early bile leaks after orthotropic liver trans-
plications in the T-tube group (33%), as compared with patients plantation are reported to occur at the T-tube insertion site. They
transplanted without T-tube (15.5%). Sixty percent of complications may be related to elective or incidental T-tube removal (Fig. 2).
in the T-tube group were related to the T-tube, with cholangitis Complications were reported in up to 33% of all T-tube remov-
being the most prominent complication.9,10 The complication rates als.13–16 The relative risk of the use of stent or T-tube splinting was
for cholangitis were extraordinarily high in this trial similar to a calculated to be 2.1 in a small cohort of patients.17 Early T-tube
third randomized trial evaluating 107 patients.11 insertion site leaks may reflect relative downstream obstruction or
FIGURE 2. Laboratory values demon-
strate comparable liver function
within the first month in both groups.
© 2009 Lippincott Williams & Wilkins www.annalsofsurgery.com | 769

5. Weiss et al Annals of Surgery • Volume 250, Number 5, November 2009
papillary dysfunction. They usually respond to unclamping of the results we conclude that the insertion of T-tubes is a save and
T-tube, placement of a transhepatic cholangiography-guided drain- cost-effective measure, which has couple diagnostic and thera-
age, Yamakawa-drainage, or endoscopic sphincterotomy, or stent- peutic advantages and should be considered for the biliary anas-
ing.18 The incidence may be lowered by tunneling the T-drain tomosis in liver transplantation.
through the mesocolon on its way outside the abdominal cavity.
Other centers have proposed a modified technique of T-tube re-
moval, using the T-tube itself as a counter-drain under fluoroscopy REFERENCES
guidance.19 An impact of T-tube removal earlier than the usual 1. Verran DJ, Asfar SK, Ghent CN, et al. Biliary reconstruction without T-tubes
period of 6 weeks to 3 months after orthotropic liver transplantation or stents in liver transplantation: report of 502 consecutive cases. Liver
Transpl Surg. 1997;3:365–373.
has not yet been confirmed.15 This uncertainty about the early
2. Rabkin JM, Orloff SL, Reed MH, et al. Biliary tract complications of
removal could be overcome by endoscopic biliary stenting, which side-to-side without T-tube versus end-to-end with or without T-tube after
was shown to facilitate safe and early removal of a T-Tube in liver choledochocholedochostomy in liver transplant recipients. Transplantation.
transplant patients.20 1998;65:193–199.
However we could confirm some of these problems in our 3. Verdonk RC, Buis CI, Porte RJ, et al. Anastomotic biliary strictures after liver
series. We diagnosed leakages at the insertion site; these problems transplantation: causes and consequences. Liver Transpl. 2006;12:726 –735.
could be resolved conservatively with a sufficient decompression of 4. Welling TH, Heidt DG, Englesbe MJ, et al. Biliary complications following
the bile duct system. We did not experience any complications after liver transplantation in the model for end-stage liver disease era: effect of
donor, recipient, and technical factors. Liver Transpl. 2008;14:73– 80.
removal of T-tubes in this series, which may be due to the fact that
5. Greif F, Bronsther OL, Van Thiel DH, et al. The incidence, timing, and
we have an extensive experience in handling these drainages, which management of biliary complications after orthotopic liver transplantation.
are an integral part of our transplantation and hepatobiliary surgery. Ann Surg. 1994;219:40 – 45.
In contrast we observed in the group without T-tubes more severe 6. Stratta RJ, Wood RP, Langnas AN, et al. Diagnosis and treatment of biliary
complications including pancreatitis and reoperations due to in- tract complications after orthotopic liver transplantation. Surgery. 1989;106:
creased total number of therapeutic and diagnostic invasive inter- 675– 683.
ventions. This finding is in accordance with a prospective trial 7. Ludwig J, Wiesner RH, Batts KP, et al. Acute vanishing bile duct syndrome
(acute irreversible rejection) after orthotopic liver transplantation. Hepatol-
including 98 patients, which reported about similar results. Interest- ogy. 1987;7:476 – 483.
ingly in this trial comparable to our study the number of surgeons 8. Wojcicki M, Milkiewicz P, Silva M. Biliary tract complication after liver
performing the transplantation was limited, providing adequate ex- transplantation: a review. Dig Surg. 2008;25:245–257.
perience with the handling of T-tubes. 9. Vougas V, Rela M, Gane E, et al. A prospective randomized trial of bile duct
In our study we could demonstrate that the rate of strictures and reconstruction at liver transplantation: T tube or no T tube? Transpl Int.
ischemic type biliary lesions after 1-year follow-up was although not 1996;9:392–395.
significantly, higher in the group without T-tubes. Similar results are 10. Scatton O, Meunier B, Cherqui D, et al. Randomized trial of choledochocho-
ledochostomy with or without a T-tube in orthotopic liver transplantation.
reported in the early period in other reports, but in most publications are Ann Surg. 2001;233:432– 437.
not considered any more over the long-term.21 As a consequence, the 11. Amador A, Charco R, Marti J, et al. Cost/efficacy clinical trial about the use
frequency of diagnostic and therapeutic interventions were increased in of T-tube in cadaveric donor liver transplant: preliminary results. Transplant
group 2 significantly enhancing the patients risk for further mortality Proc. 2005;37:1129 –1130.
and complications, which was confirmed in our trial. 12. Kusano T, Randall HB, Roberts JP, et al. The use of stents for duct-to-duct
Several groups analyzed the economic impact of the use of anastomoses of biliary reconstruction in orthotopic liver transplantation.
Hepatogastroenterology. 2005;52:695– 699.
T-tubes and an analysis of cost-effectiveness, respectively.21 In
13. Ostroff JW, Roberts JP, Gordon RL, et al. The management of T-tube leaks
these analysis the application of T-tubes resulted in significantly in orthotopic liver transplant recipients with endoscopically placed nasobili-
higher complication rates (32.9% vs. 15.5% without T-tube), how- ary catheters. Transplantation. 1990;49:922–924.
ever, complication related costs were not significantly higher in one 14. O’Connor TP, Lewis D, Jenkins RL. Biliary tract complications after liver
of the studies emphasizing the fact that the complications were transplantation. Arch Surg. 1995;130:312–317.
minor issues in this group. In out trial hospital stay, radiologic 15. Shuhart MC, Kowdley KV, McVicar JP, et al. Predictors of bile leaks after
studies, and cost of hospital resources were higher among the T-tube removal in orthotopic liver transplant recipients. Liver Transpl Surg.
patients without T-tubes, as the complications in this group were 1998;4:62–70.
more severe and cost intensive followed by complex diagnostic and 16. Grande L, Perez-Castilla A, Matus D, et al. Routine use of the T-tube in the
´
biliary reconstruction of liver transplantation: is it worthwhile? Transplant
therapeutic measures.22 Proc. 1999;31:2396 –2397.
Since several groups apparently have felt the need for secur- 17. Qian YB, Liu CL, Lo CM, et al. Risk factors for biliary complications after
ing the biliary anastomoses by catheters and stents, other alternatives liver transplantation. Arch Surg. 2004;139:1101–1105.
have been proposed, such as transcystic indwelling catheters. It was 18. Osorio RW, Freise CE, Stock PG, et al. Nonoperative management of biliary
concluded that both techniques were equally effective in obtaining a leaks after orthotopic liver transplantation. Transplantation. 1993;55:1074 –
satisfactory postoperative cholangiogram, however, the transcystic 1077.
catheter technique allowed a significantly earlier withdrawal with 19. Urbani L, Campatelli A, Romagnoli J, et al. T-tube removal after liver
transplantation: a new technique that reduces biliary complications. Trans-
fewer complications compared with the T-tube technique and maybe plantation. 2002;74:410 – 415.
therefore be considered as a potential alternative to T-tubes.11,23 20. Rolles K, Fusai G, Rolando N, et al. Endoscopic biliary stenting facilitates
This prospective randomized monocentric trial proofs that the safe and early removal of T-tube in liver transplant patients. Minerva Chir.
rate of severe complications after side-to-side choledochochole- 2005;60:31–35.
dochostomy is higher in the group without T-tubes. This is in 21. Verdonk RC, Buis CI, Porte RJ, et al. Biliary complications after liver
contrast to former published results. However the usage of a T-tube transplantation: a review. Scand J Gastroenterol Suppl. 2006;5:89 –101.
prevented and decreased the need of postoperative therapeutic and 22. Shimoda M, Saab S, Morrisey M, et al. A cost-effectiveness analysis of
biliary anastomosis with or without T-tube after orthotopic liver transplanta-
diagnostic measures and their inherent risks. The number of invasive tion. Am J Transplant. 2001;1:157–161.
therapeutic procedures was increased in the group without T-tube 23. Innocenti F, Hepp J, Humeres R, et al. Transcystic cholangiogram access via
within the first year leading to further complications, the long- rubber band with early withdrawal after liver transplantation: a safe tech-
term results are in favor of the usage of T-tubes. Considering our nique. Transplant Proc. 2004;36:1681–1682.
770 | www.annalsofsurgery.com © 2009 Lippincott Williams & Wilkins

6. Annals of Surgery • Volume 250, Number 5, November 2009 Biliary Complications After Liver Transplantation
I think that in such studies the policy of the ELTR registry should be
Discussions followed, which means that you should divide complications occur-
ring within 3 months, what we call early complications, from
PROFESSOR J. LERUT (BRUSSELS, BELGIUM): This article is complications occurring after 3 months, what we call late compli-
another that deals with a problem that has not been solved despite all cations. Such methodology could completely change the conclusions
of the progress made in transplantation from 1963 to 2009; that is, of this study.
the issue of biliary tract complications. If we look back at the first The incidence of ischemic tract biliary lesions was only
Pittsburgh publication in 1983 regarding 400 patients, the incidence around 5%, which is a very low figure. Do you routinely check intra
of biliary tract complications was 18% and the incidence is still 18% and extra bile ducts in all your patients not only after 7 days or 6
to 20%. Many articles have been published in relation to this subject, weeks when you take out the tube, but also after 3, 6, and 12
but only 2 were published in peer reviewed journals: the King’s months? We, in our vast Brussels’ experience, showed that early
College Group mentioned there was no difference at all between biliary tract controls are insufficient; you can indeed have many
T-tubing and not, and the French multicentre study published in the patients with a completely normal intra and extrahepatic bile duct
Annals of Surgery in 2001 was in favor of performing the bile duct system at 1, 2, 3, 4, 5, and 6 weeks; if you conduct a percutaneous
anastomosis without T-tube because there were more infections. transhepatic cholangiogram at 6 or 12 months they may present a
This is in contrast to your observation. You indeed had fewer completely asymptomatic biliary tract destruction with even com-
infections in the group with T-tube, but these patients were also pletely normal cholestatic and noncholestatic enzymes.
transplanted in another period. I think it is difficult to compare today
what happened 10 years ago because perioperative, anti-infectious JOHANN PRATSCHKE (BERLIN, GERMANY): First I want to com-
and especially immunosuppressive care has completely changed, so ment on your question as to why we included the pancreatitis and
it is to be expected that the infection rate has also changed. You try endoscopic interventions as complication in our trial. I do not agree
to address the biliary tract problem in a large single center study. In with your statement that this was not a consequence of the technical
the mini abstract you state that the use of a T-tube is a safe and aspects. After diagnosis of a biliary leakage the patient usually needs
excellent tool to control liver function and biliary tract reconstruc- frequent endoscopies. I think these complications must be taken into
tion. However, as liver function can be easily monitored by lactic account; these kinds of problems could potentially be avoided by
acid and factor V determination, I do not think we need bile using a T-tube. Naturally, we also have insufficiencies in the T-tube
production for monitoring graft function. In the abstract of the article group, but these leakages usually do not need an endoscopic inter-
you added that the patients with a T-tube experienced a significantly
vention because the bile duct is decompressed. You can wait and
lower complication rate. These conclusions deserve some comment.
see, and in most instances you observe a spontaneous resolution of
The significant differences in favor of the T-tube were obtained in
the problem. I think this justifies the inclusion of these complications
your study by including in the analysis not only “pure ” technical
in the study as they are consequences of the technique of bile duct
complications such as stenosis or leakage, but also by adding
anastomoses. Your next question asked when we consider pancre-
complications with a very different physiopathology to the analysis;
complications such as papillary stenosis unrelated to the tube inser- atitis as a complication. Of course after an endoscopic intervention
tion, ischemic type biliary tract lesions at only 1 year of follow-up nearly 100% of patients react with moderate increases in lipase and
(which is much too short), and, last but not least, an important amylase levels. These elevations were not considered as pancreatitis,
number of pancreatic and cholangitic episodes related not only to however, if the values are per definition 3 times over baseline, after
early, but also to later endoscopic bile duct imaging. Both latter 5 days pancreatic alterations were still proven by CT scan, it was
complications represent respectively 24 of 53 (45%) and 9 of 30 qualified as pancreatitis. Thus, to make the story short, if we had
(30%) complications in the non-T-tube and in the T-tube groups. after 5 days still significantly elevated lipase and amylase levels we
These complications are responsible for your significant P of 0.005, performed a CT scan to radiologically prove the diagnosis. Again, I
but if you exclude them your P value is reduced to a less significant would suggest that this severe complication is justifiably included. I
value of 0.01. This means that if one excludes both complications agree with you that we had a very low rate of the ischemic type of
from the analysis, the differences between the 2 patient groups are biliary lesions, and I also agree with you that it is difficult to make
faded in relation to the real reasons for using or not using a T-tube, conclusions from the findings after 6 weeks to long-term results after
this means avoiding leakage and especially stenosis. We know from 1 year. We see all our patients on a regular basis in the hospital over
the usual biliary tract surgery that a T-tube avoids scarring of the the long term. We do not perform routine endoscopy or radiologic
suture thereby delaying stenosis. imaging. The frequency of an ischemic type biliary lesion was low
I have some questions for you. Why did you really include in our series, I am sure we would diagnose more of these alterations
pancreatitis and cholangitis in this analysis especially when taking if we assessed them through a specific diagnostic workup. Our
into account that the French multicentre study showed exactly the assessment of this entity was based on elevated cholestatic param-
opposite in relation to infections? Second, what was your definition eters and following radiologic imaging, and if necessary endoscopy.
of periendoscopic pancreatitis. You may indeed have a marked I agree with you on the statement that if we look more carefully at
biochemical elevation of amylase, but this does not mean the patient the details after 1 year we will diagnose more ITBL. However the
is sick; you should also indicate when these complications occurred. distribution was nearly equal in both groups.
© 2009 Lippincott Williams & Wilkins www.annalsofsurgery.com | 771