Yamamoto S, Sato Y, Oya H, Nakatsuka H, Kobayashi T, Hara Y, Watanabe T, Kurosaki I, Hatakeyama K. Risk factors and prevention of biliary anastomotic complications in adult living donor liver transplantation. World J Gastroenterol 2007; 13(31): 4236-4241 [PMID: 17696254 DOI: 10.3748/wjg.v13.i31.4236]
Corresponding Author of This Article
Satoshi Yamamoto, MD, Division of Digestive and General Surgery, Niigata University Graduate School of Medical and Dental Sciences, Asahimachi-dori 1-757, Niigata 951-8510, Japan. sugarman@med.niigata-u.ac.jp
Article-Type of This Article
Rapid Communication
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Satoshi Yamamoto, Yoshinobu Sato, Hiroshi Oya, Hideki Nakatsuka, Takashi Kobayashi, Yoshiaki Hara, Takaoki Watanabe, Isao Kurosaki, Katsuyoshi Hatakeyama, Division of Digestive and General Surgery, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan
ORCID number: $[AuthorORCIDs]
Author contributions: All authors contributed equally to the work.
Correspondence to: Satoshi Yamamoto, MD, Division of Digestive and General Surgery, Niigata University Graduate School of Medical and Dental Sciences, Asahimachi-dori 1-757, Niigata 951-8510, Japan. sugarman@med.niigata-u.ac.jp
Telephone: +81-25-2272228 Fax: +81-25-2270779
Received: May 22, 2007 Revised: June 3, 2007 Accepted: June 9, 2007 Published online: August 21, 2007
Abstract
AIM: To evaluate risk factors of biliary anastomotic complications (BACs) and outcomes according to type of biliary reconstruction.
METHODS: A total of 33 consecutive adult living donor liver transplantation (LDLT) were reviewed, 17 of which had undergone Duct-to-Duct anastomosis (D-D). The remaining 16 patients received Roux-en-Y anastomosis (R-Y). The perioperative factors, such as the type of graft and the number of graft bile ducts, were analyzed retrospectively.
RESULTS: The overall incidence of BACs was 39.4%. The incidence of BACs was significantly higher in the patients with than without neoadjuvant chemotherapy (71.4% vs 10%, P = 0.050). There was no significant difference in the incidence of biliary leakage in patients with D-D vs those with R-Y. The incidence of biliary strictures following the healing of biliary leakage was significantly higher in D-D (60%) than in R-Y (0%) (P = 0.026). However, the incidence of BACs related bacteremia was significantly higher in R-Y than in D-D (71.4% vs 0%, P = 0.008). In D-D, use of T-tube stent remarkably reduced the incidence of BACs, compared with straight tube stent (0% vs 50%, P = 0.049).
CONCLUSION: Our experience showed an increase of BACs related bacteremia in the patients with R-Y. Therefore, D-D might be a preferred biliary reconstruction. However, the surgical refinement of D-D should be required because of the high incidence of biliary strictures. Use of the T-tube stent might lead to a significant reduction of BACs in D-D.
Citation: Yamamoto S, Sato Y, Oya H, Nakatsuka H, Kobayashi T, Hara Y, Watanabe T, Kurosaki I, Hatakeyama K. Risk factors and prevention of biliary anastomotic complications in adult living donor liver transplantation. World J Gastroenterol 2007; 13(31): 4236-4241
Since the early period of orthotopic liver transplantation (OLT), biliary reconstruction has been characterized as the technical Achilles’ heel of OLT[1-3]. This holds true also for living donor liver transplantation (LDLT)[4,5]. At present, standard biliary reconstruction in OLT is a choledocho-choledochostomy. On the other hand, Roux-en-Y anastomosis (R-Y) is the preferred technique for biliary reconstruction in LDLT or reduced liver transplantation.
Recently, Duct-to-Duct anastomosis (D-D) has been widely performed for biliary reconstruction even in adult LDLT[6-8], because of its advantages, for example concerning the prevention of reflux cholangitis and the reduced duration of surgery[9,10]. In addition, D-D provides improved access to the biliary tract for endoscopic retrograde cholangiography. However, in several early reports, the incidence of biliary anastomotic complications (BACs), especially biliary strictures, were higher in patients having received the D-D than in those in whom R-Y had been performed[11-14]. Therefore, there is no consensus regarding biliary reconstruction in LDLT yet. In this study, we evaluated whether there was a difference in the incidence of BACs according to the type of biliary reconstruction in adult LDLT.
MATERIALS AND METHODS
The adult LDLT programme started at our institution in March 1999. The period from March 1999 to February 2004 was chosen for our analysis, since it provided a follow-up longer than three years. During this period, 35 adult patients underwent LDLT at our institute. During roughly the first half of our study, we mainly performed R-Y for biliary reconstruction. However, since middle 2001, we switched from R-Y to D-D because of its advantages. Through the end of the study period, 18 patients have undergone the D-D and 17 patients have undergone the R-Y. Two patients, who died within one month after surgery, were excluded from the study. The causes of the death were persistent bacteremia (existing prior to operation) and cerebral hemorrhage. The demographic characteristics of the 33 consecutive recipients included in this study are shown in Table 1. We analyzed etiologic factors affecting BACs and evaluated the differences in the incidence of BACs between the patients having received D-D and those treated with a R-Y.
Healthy individuals aged 19 to 62 years were eligible donors. The selection of the left versus right liver lobe was based on donor graft liver volume as measured through computer-assisted tomography scans (CT scans). Essentially, whichever lobe had a graft-to-recipient body weight ratio (GRWR) greater than 0.7% was used as the graft liver. However, when the right liver lobe exceeded 65% of the donor’s whole liver volume, the left liver lobe was occasionally selected with fully informed consent even if the GRWR was less than 0.7%. All donor patients preoperatively underwent DIC-CT scanning to measure graft liver volume and to evaluate the bile duct.
Donor lobectomy was performed as described previously in detail[15-17]. After cholecystectomy, cystic duct cholangiography was performed to confirm the transection line. Hilar dissection was then performed. The left or right hepatic ducts were transected sharply and parenchymal transection was performed. The peribiliary plexus was preserved as much as possible. After systemic heparinization with 1000 units of heparin, the graft was flushed via the portal vein with three times the graft weight of histidine-tryptophan-ketoglutarate (HTK) at 4°C.
To decompress the biliary tract, C-tubes were routinely positioned in the common bile duct via the cystic duct in all donor cases[18].
Recipient operations
Recipient total hepatectomy was performed with preservation of the inferior vena cava (IVC). After anastomosis of the hepatic vein, significant large accessory veins (> 5 mm) were also anastomosed to the sidewall of the IVC[19]. The donor portal vein was anastomosed to the recipient’s main portal vein. After reperfusion of the graft, arterial anastomosis was completed through surgical microscopy between the donor hepatic artery and the recipient right, left or proper hepatic artery in most cases.
Biliary reconstruction
R-Y was performed in 16 patients, 6 with a right liver lobe and 10 with a left liver lobe. The bile duct was anastomosed to the Roux-en-Y limb of the jejunum by using a 6-0 PDS suture in an interrupted fashion. The 4-, 6-, or 8-Fr straight tubes were routinely inserted and introduced via the R-Y limb of the jejunum as a stent tube.
Seventeen patients underwent D-D. For cases with D-D anastomosis, the recipient bile duct was dissected at the hilar plate with as much surrounding tissue attached as possible. An end-to-end anastomosis was performed with the placement of an external stent by using a 6-0 PDS suture in an interrupted fashion in cases whose underlying hepatobiliary disease allowed us to use this technique. Initially, a 4-, 6-, or 8-Fr straight tube was routinely positioned in the intrahepatic bile duct via the remnant cystic duct or the sidewall of the common bile duct. We have recently begun using a 9- or 12-Fr T-tube for an external stent to reduce BACs. We remove the T-tube between 3 and 6 mon after the LDLT.
Diagnosis of posttransplant complications
When we found a discharge of bilirubin occurring through the drains, we confirmed anastomotic leakages by performing cholangiography using external stent tubes.
Biliary stricture is primarily suspected with an increase of liver function tests or a presence of jaundice. Confirmation of intrahepatic bile duct dilatation using abdominal ultrasonography or CT scan was possible in most of the cases.
The patient was diagnosed with bacteremia when the presence of bacteria was showed in the patient’s blood. We concluded that the bacteremia was associated with BACs wnever the same bacterial species were found in the patient’ bile and blood.
Statistical analysis
Values were given as means ± SD. Data was analyzed by the chi-square test and the Mann-Whitney U-test. Patient survivals after LDLT were determined by the Kaplan-Meier method and compared among groups using Log-rank test. A difference was considered statistically significant when probability was less than 0.05.
RESULTS
Rates of patient survival
The mean follow-up period of the 33 patients was 47.2 ±28.9 mo. Eight study patients died, and the overall survival rate of the 33 patients was 75.8%. The causes of death were systemic infection resulting in liver graft failure at 3 mo, 5 mo, or 7 mo (n = 3); heart failure following amyloidosis at 7 mo (n = 1); and the recurrence of hepatocellular carcinoma (HCC) at 6 mo, 8 mo, or 10 mo (n = 3). In two of three patients died due to systemic infection, the infection contributed to repeated reflux cholangitis after biliary leakage. The cumulative patient survival rate tended to be lower in patients without than with BACs (Figure 1).
Figure 1 Cumulative patient’s survival with or without biliary anastomotic complications (BACs).
The cumulative patient survival rate tended to be lower in the patients with BACs (n = 13) than without BACs (n = 20) (P = 0.071).
Rates and risk factors of biliary anastomotic complications
The overall incidence of BACs was 39.4% (13/33). Biliary leakage developed in 11 patients, while biliary stricture occurred in 2 patients. We analyzed the risk factors of BACs by comparing patients with and without BACs (Table 2). The patients with neoadjuvant chemotherapy (NAC) for hepatocellular carcinoma (HCC) showed significantly higher incidence of BACs than those without NAC (71.4% vs 10%, P = 0.050, Table 2). Seven patients underwent NAC, of which 5 experienced BACs (71.4%). NAC was defined as systemic imuunochemotherapy or transarterial chemoembolization (TACE) performed within three months before LDLT.
Table 2 Preoperative risk factor for biliary anastomotic complications in adult living donor liver transplantation.
Biliary anastomotic complications
Absent (n = 20)
Present (n = 13)
P value
Age (yr)
50.4 ± 12.8
46.2 ± 16.4
NS
Gender
NS
Male
10
7
Female
10
6
CTP
10.4 ± 3.2
10.0 ± 3.2
NS
MELD
18.5 ± 8.0
16.8 ± 7.4
NS
ABO matching
NS
Identical
15
10
Compatible
2
3
Incompatible
3
0
NAC
0.05
Yes
2
5
No
18
2
Donor age (yr)
38.4 ± 11.6
33.8 ± 13.5
NS
Cold ischemia time, hepatic artery thrombosis or CMV infection, which has been identified as a cause of BACs in the previous reports, was not observed in any recipients in this series (Table 3, Table 4), neither was there an ABO-incompatible case (Table 2). Several recent reports have shown that multiple biliary orifices, which present a further hurdle in biliary reconstruction, significantly increased the incidence of BACs[14]. However, in our study, although the number of BACs was increased in cases multiple bile ducts, the difference failed to reach statistical significance due to the low number of cases. The incidence of posttransplant bacteremia was significantly higher in patients with than without BACs (53.8% vs 10%, P = 0.006, Table 4). There was no significant difference in any of the other risk factors between patients with and without BACs.
Table 3 Intraoperative risk factor for biliary anastomotic complications in adult living donor liver transplantation.
Biliary anastomotic complications
Absent (n = 20)
Present (n = 13)
P value
WIT (min)
47.8 ± 9.1
46.0 ± 7.7
NS
CIT (min)
61.7 ± 52.9
47.1 ± 32.9
NS
Post PVP (mmHg)
163.2 ± 40.7
182.3 ± 42.5
NS
Graft
NS
Right
4
4
Left
16
9
GRWR (%)
0.834 ± 0.211
0.843 ± 0.261
NS
Type of reconstruction
NS
Duct-to-Duct
11
6
Roux-en-Y
9
7
Number of bile duct
NS
Single
16
8
Multiple
4
5
Table 4 Postoperative risk factor for biliary anastomotic complications in adult living donor liver transplantation.
Biliary anastomotic complications
Absent (n = 20)
Present (n = 13)
P value
HAT
NS
Yes
0
0
No
20
13
ACR
NS
Yes
10
6
No
10
7
CMV
NS
Yes
10
6
No
10
7
Bacteremia
0.006
Yes
2
7
No
18
6
Analysis of biliary anastomotic complications according to the type of biliary reconstruction
In the 33 adult LDLT patients, there were 13 BACs, including 11 cases of leakage (33.3%) and 2 strictures (6.1%) (Table 5). Biliary leakage was observed in 6 (37.5%) of 16 R-Y patients and in 5 (29.4%) of 17 D-D patients. Biliary strictures developed in one patient in each group. Both of the patients had one anastomosis with two bile ducts. There were no significant differences in the incidence of BACs between R-Y and D-D. In all patients who experienced biliary leakage, the leakage was halted by continuous percutaneous biliary drainage or endoscopic nasobiliary drainage. None of the patients required reoperation for the leakage. However, in 3 of the 5 patients who experienced biliary leakage in D-D, a biliary stricture developed after the leakage had healed. To manage the stricture, all 3 patients required percutaneous transhepatic biliary drainage (PTCD), followed by tube stenting. One patient required a conversion to R-Y anastomosis because of the complete obstruction of the posterior branch. On the other hand, no R-Y patient experienced biliary stricture following the healing of anastomotic leakage (60% vs 0%, P = 0.026, Table 5).
Table 5 Incidence of biliary anastomotic complications according to type of biliary reconstructions.
Stricture
Leakage
Leakage (-)
Leakage (+)
Duct-to-Duct (n = 17)
5
1
3
Roux-en-Y (n = 16)
6
1
0
Incidence of bacteremia associated with biliary anastomotic complications according to type of biliary reconstructions
5 of 7 R-Y patients with BACs experienced bacteremia associated with BACs (71.4%), while none of 6 D-D patients with BACs had bacteremia associated with BACs (0%). The incidence of BACs related bacteremia was significantly higher in R-Y patients than in D-D patients among the patients experienced BACs (P = 0.008, Table 6)
Table 6 Incidence of bacteremia associated with biliary anastomotic complications according to type of biliary reconstructions.
Absent
Present
P value
Duct-to-Duct (n = 6)
6
0
Roux-en Y (n = 7)
2
5
0.008
Incidence rate of biliary anastomotic complications according to the type of external stent tube
The 4- or 6-Fr straight tube was placed via the stoma in 12 D-D patients (Figure 2A). The remaining 5 patients had 9- or 12-Fr T-tube as an external stent (Figure 2B). There were no complications in the biliary tract in 5 LDLT patients with T-tube drainage, while BACs occurred in 6 of 12 patients with straight tube drainage (P = 0.049, Table 7).The stent tube was usually left in place for at least 3 mo after LDLT and then removed after confirming the integrity of the bile duct by a cholangiogram. There was no biliary complication associated with the insertion or removal of T-tube, such as bile leakage at the T-tube insertion and biliary peritonitis after T-tube removal.
Figure 2 External stent tube in Duct-to-Duct anastomosis.
A: 6 Fr straight tube via the sidewall of the common bile duct; B: 9 Fr T-tube.
Table 7 Incidence of biliary anastomotic complications according to type of external stent tubes in Duct-to-Duct anastomoses.
Absent
Present
P value
Straight tube (n = 12)
6
6
T tube (n = 5)
5
0
0.049
DISCUSSION
Although surgical, technological, and immunological treatments have advanced greatly in the field of liver transplantation, biliary anastomotic complications remain an important cause of morbidity and mortality, as confirmed by our series.
In our study, there was no difference in the incidence of BACs between the D-D and the R-Y. However, the incidence of biliary strictures following the healing of biliary leakage was significantly higher in the D-D (60%) than in the R-Y (0%). This finding emphasizes that ischemia is an important cause of BACs. Although the Roux-en-Y limb of the jejunum receives an adequate arterial blood flow, ischemia might occur easily in the recipient’s bile duct because of lysis of porta hepatis. Therefore, in comparison with R-Y, D-D has to be monitored closely in order to preserve the peribiliary plexus of the recipient’s left, right, or common hepatic duct as well as the donor’s right or left hepatic duct. The high incidence of biliary strictures in D-D probably might originate in a little longer stricture caused by ischemic change in both sides of the stoma.
However, D-D has several advantages over R-Y. First, D-D can reduce the duration of surgery. Second, D-D allows easier access for endoscopic treatments of the biliary tract and may reduce reflux cholangitis because of preservation the normal physiologic sphincter mechanism. Third, D-D can be converted to R-Y in cases where biliary strictures cannot be resolved by percutaneous transhepatic biliary anastomotic dilatations. Moreover, in our study, the incidence of bacteremia associated with BACs was significantly lower in the D-D than in the R-Y cases. Therefore, D-D is our technique of choice for biliary reconstruction. However, surgical refinement of D-D anastomosis should be performed because of the high incidence of biliary strictures, as reported previously[20]. Early in our study period, a 4- or 6-Fr straight tube was used as an external stent in D-D, as in R-Y. However, stent tubes of these sizes often caused obstruction; several patients who had been given the narrow hepatic duct experienced biliary congestion or cholangitis after the external stent tube was clamped. Since January 2003, we started using a T-tube of 9- or 12-Fr in diameter, which suits the hepatic duct exactly, as an external stent. Previously, Randall demonstrated that the use of the T-tube did not reduce the biliary complication rate in liver transplantation[21]. However, in LDLT, the D-D anastomosis is a hepaticohepaticostomy, in which a narrow stoma and peristomal ischemia may easily occur. As reported previously, the incidence of BACs should be higher in LDLT than in whole liver transplantation. Actually, our experience showed that there were no such complications in the biliary tract in 5 LDLT patients with T-tube drainage, while patients with straight tube drainage experienced high incidence of BACs. Hashimoto also reported that use of T tube in LDLT might reduce the incidence of biliary stricture and can be an effective therapeutic option when endoscopic treatment is unsuccessful[30]. Therefore, we expect that the use of the T-tube may reduce the incidence of BACs, especially biliary stictures in LDLT. More experience is needed to confirm this.
Heffron and Reichert reported a high incidence of BACs in reduced-size liver transplantation or LDLT in comparison with cadaveric whole liver transplantation[3,23]. Prolonged cold ischemia, chronic rejection and cytomegalovirus infection have been proposed as important risk factors for BACs after liver transplantation[24,25]. However, in LDLT, there have not been so many reports regarding the incidence of BACs.
We analyzed the risk factors between patients with BACs and those without. The incidence of BACs was significantly higher in the patients with NAC. Previously, Kim and Kemeny reported ischemic change in the bile duct after TACE[26,27]. Ischemia is the most important cause of BACs as described above. It is possible that LDLT patients with HCC have suffered from ischemic bile duct injury following NAC and thus are more likely to undergo BACs after LDLT. In the future, a larger group of patients should be studied to verify this hypothesis.
In this series, the overall incidence of BACs was 39.4%. It is true that the incidence rate seems to be slightly higher than that in previous reports (18.2%-43%)[4,5,12,28,29]. However, these reports described little of a correlation between BACs and NAC. The high incidence of BACs in our study may be due to the inclusion of many patients with NAC.
Recently, a few authors found that multiple bile ducts in the graft are strongly correlated with a high risk of BACs[14,30,31]. Multiple bile ducts showed an increase of BACs also in our study. Among the nine patients with multiple bile ducts, five had BACs (55.6%), whereas among 25 patients with single bile duct, BACs occurred only in eight (32%). This difference was not statistically significant due to the low number of cases, and needs to be confirmed in the future by larger studies.
Although there was no significant difference in the incidence of BACs between the left-lobe and right-lobe patients, the patients with the right-lobe graft clearly had increased BACs, compred with those with the left-lobe (50% vs 36%). Marcos has previously reported that the incidence of multiple bile ducts or anatomical variations is very high in right-lobe graft. In our series, the incidence of multiple bile ducts was significantly higher in the right lobe graft than in the left lobe graft (60% vs 12%, P = 0.014, data not shown). The increased incidence of BACs in the right lobe graft might reflect the high incidence of BACs in patients with multiple bile ducts.
In summary, our experience showed an increase of bacteremia associated with BACs in the patients having received R-Y. Therefore, D-D might be a preferred and safer biliary reconstruction than R-Y. However, the surgical refinement of D-D should be required because of the high incidence of biliary strictures. Use of the T-tube stent might lead to a significant reduction of BACs in D-D.
Stratta RJ, Wood RP, Langnas AN, Hollins RR, Bruder KJ, Donovan JP, Burnett DA, Lieberman RP, Lund GB, Pillen TJ. Diagnosis and treatment of biliary tract complications after orthotopic liver transplantation.Surgery. 1989;106:675-683; discussion 683-684.
[PubMed] [DOI][Cited in This Article: ]
Miller CM, Gondolesi GE, Florman S, Matsumoto C, Muñoz L, Yoshizumi T, Artis T, Fishbein TM, Sheiner PA, Kim-Schluger L. One hundred nine living donor liver transplants in adults and children: a single-center experience.Ann Surg. 2001;234:301-311; discussion 311-312.
[PubMed] [DOI][Cited in This Article: ]
Broelsch CE, Burdelski M, Rogiers X, Gundlach M, Knoefel WT, Langwieler T, Fischer L, Latta A, Hellwege H, Schulte FJ. Living donor for liver transplantation.Hepatology. 1994;20:49S-55S.
[PubMed] [DOI][Cited in This Article: ]
Sato Y, Kurosaki I, Yamamoto S, Nakatsuka H, Oya H, Shirai Y, Tanaka K, Hatakeyama K. Postoperative management for donor safety in living related donor liver transplantation.Hepatogastroenterology. 2003;50:196-200.
[PubMed] [DOI][Cited in This Article: ]
Kawachi S, Shimazu M, Wakabayashi G, Hoshino K, Tanabe M, Yoshida M, Morikawa Y, Kitajima M. Biliary complications in adult living donor liver transplantation with duct-to-duct hepaticocholedochostomy or Roux-en-Y hepaticojejunostomy biliary reconstruction.Surgery. 2002;132:48-56.
[PubMed] [DOI][Cited in This Article: ][Cited by in Crossref: 84][Cited by in F6Publishing: 92][Article Influence: 4.2][Reference Citation Analysis (0)]
