LIVER TRANSPLANTATION 20:893–903, 2014

ORIGINAL ARTICLE

Risk Factors and Surgical Management of Anastomotic Biliary Complications After Pediatric Liver Transplantation  e,1 Tom Darius,1* Jairo Rivera,1,4* Fabio Fusaro,1 Quirino Lai,2 Catherine de Magne 1 1 3 Christophe Bourdeaux, Magdalena Janssen, Philippe Clapuyt, and Raymond Reding1 1 Pediatric Surgery and Transplant Unit, 2Starzl Unit of Abdominal Transplantation, and 3Pediatric Radiology Department, University Clinics Saint-Luc, Universite Catholique de Louvain, Brussels, Belgium; and 4 Transplant Unit, Fundacion Cardioinfantil IC, Bogota, Colombia

Biliary complications (BCs) still remain the Achilles heel of liver transplantation (LT) with an overall incidence of 10% to 35% in pediatric series. We hypothesized that (1) the use of alternative techniques (reduced size, split, and living donor grafts) in pediatric LT may contribute to an increased incidence of BCs, and (2) surgery as a first treatment option for anastomotic BCs could allow a definitive cure for the majority of these patients. Four hundred twenty-nine primary pediatric LT procedures, including 88, 91, 47, and 203 whole, reduced size, split, and living donor grafts, respectively, that were performed between July 1993 and November 2010 were retrospectively reviewed. Demographic and surgical variables were analyzed, and their respective impact on BCs was studied with univariate and multivariate analyses. The modalities of BC management were also reviewed. The 1- and 5-year patient survival rates were 94% and 90%, 89% and 85%, 94% and 89%, and 98% and 94% for whole, reduced size, split, and living donor liver grafts, respectively. The overall incidence of BCs was 23% (n 5 98). Sixty were anastomotic complications [47 strictures (78%) and 13 fistulas (22%)]. The graft type was not found to be an independent risk factor for the development of BCs. According to a multivariate analysis, only hepatic artery thrombosis and acute rejection increased the risk of anastomotic BCs (P < 0.001 and P 5 0.003, respectively). Anastomotic BCs were managed primarily with surgical repair in 59 of 60 cases with a primary patency rate of 80% (n 5 47). These results suggest that (1) most of the BCs were anastomotic complications not influenced by the type of graft, and (2) the surgical management of anastomotic BCs may constitute the first and best therapeutic option. Liver Transpl 20:893-903, C 2014 AASLD. 2014. V Received December 10, 2013; accepted May 6, 2014.

Abbreviations: BC, biliary complication; CBD, common bile duct; CD, cystic duct; CHD, common hepatic duct; CI, confidence interval; CMV, cytomegalovirus; HTK, histidine tryptophan ketoglutarate; LHD, left hepatic duct; LT, liver transplantation; NA, not available; OR, odds ratio; RHAP, retrograde hepatic artery perfusion; RHD, right hepatic duct. The authors have no grants or other financial support or conflicts of interest to declare. Raymond Reding, Jairo Rivera, and Fabio Fusaro conceived this study. Jairo Rivera, Catherine de Magn ee, Christophe Bourdeaux, and Magdalena Janssen acquired the data. Tom Darius, Jairo Rivera, Quirino Lai, and Raymond Reding analyzed and interpreted the data. Tom Darius, Jairo Rivera, and Raymond Reding wrote the article. Quirino Lai and Philippe Clapuyt revised the article. *These authors contributed equally to this work.  Catholique de Address reprint requests to Tom Darius, M.D., Pediatric Surgery and Transplant Unit, University Clinics Saint-Luc, Universite Louvain, 10 Hippocrate Avenue, 1200 Brussels, Belgium. Telephone: 132 2 764 12 23; FAX: 132 2 762 36 80; E-mail: [email protected] DOI 10.1002/lt.23910 View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

C 2014 American Association for the Study of Liver Diseases. V

894 DARIUS ET AL.

LIVER TRANSPLANTATION, August 2014

See Editorial on Page 879

Despite what has been learned since the first instance of liver transplantation (LT) in 1963, the incidence of biliary complications (BCs) has not varied much, with the overall reported incidence still ranging from 10% to 35%.1-4 In the context of left split deceased donor grafts or left living donor grafts, microthrombosis of the peribiliary arterial plexus at the level of the left bile duct stump has been suspected of playing a crucial role in posttransplant anastomotic biliary strictures. Moreover, when a BC is diagnosed, the current therapeutic option at many centers is interventional radiology, with surgical reintervention left as a second choice. In this work, we reviewed the incidence and characteristics of BCs at our pediatric LT program over the course of 18 years. We hypothesized that (1) the use of alternative techniques (reduced size, split, and living donor grafts) may contribute to an increased incidence of BCs, and (2) surgery as a first treatment option for anastomotic BCs could allow a single-stage, long-term cure for the majority of these patients.

PATIENTS AND METHODS Recipient Demographics All primary pediatric LT procedures (patient age  18 years old) performed at the University Clinics SaintLuc (Brussels, Belgium) between July 1, 1993 and

November 1, 2010 were retrospectively reviewed. This study was approved by the institutional ethical committee. During this interval, primary pediatric LT was performed 429 times; there were 226 deceased donors and 203 living donors. The following recipient characteristics were reported: age, sex, blood group, and original liver disease. Recipient demographic data are presented in Table 1 according to the type of liver graft. Children less than 3 years old and especially children less than 1 year old were more likely to receive a technical variant graft. The use of living donors was highest among children under the age of 1 year. Biliary atresia was the most common indication for transplantation, and this was independent of the graft type.

Donor Demographics The following donor characteristics were analyzed: age, sex, blood group, type of donor [deceased donor (donation after brain death or donation after circulatory death) versus living donor], type of liver graft (whole liver, reduced size, split, or living donor), donor/recipient cytomegalovirus (CMV) match, and type of preservation solution [histidine tryptophan ketoglutarate (HTK) solution or University of Wisconsin solution]. Donor demographic data are presented in Table 2 according to the type of liver graft. University of Wisconsin solution was most frequently used for all types of grafts. Only 1 liver graft in this series originated from a donation after circulatory death donor (category 3 according to the Maastricht

TABLE 1. Recipient Demographics According to the Graft Type for a Series of 429 Pediatric LT Procedures Performed at Saint-Luc University Clinic: July 1993 to November 2010

Age (years)* Age group [n (%)] 2 mm) visualized on Doppler ultrasound, a surgical redo was planned. Accordingly, broad-spectrum antibiotics (ampicillin, 50 mg/kg, 4 times per day, and temocillin, 50 mg/kg, 2 times per day) were administered intravenously for 2 days, and they were followed by

Doppler ultrasound–guided percutaneous transhepatic cholangiography under general anesthesia. If an anastomotic stricture was evidenced, immediate surgical exploration of the biliary anastomosis was performed during the same anesthesia session. This surgical reintervention included intraoperative cholangiography through the Roux-en-Y loop to confirm the biliary anastomotic stricture. The stenotic bile duct tissue was resected in order to reach normal biliary epithelium. A new biliary anastomosis was then constructed with the Roux-en-Y loop and interrupted 7/0 resorbable monofilament stitches (PDS 7/0 Visiblack, Johnson & Johnson Medical) without the use of a biliary drain or stent. A biliary drain was not routinely left in place. The primary patency rate of the surgically treated BCs was defined as the normalization of liver enzymes, the disappearance of dilated bile ducts on ultrasound, the absence of repeated biliary surgery or interventional radiology, and the absence of a negative impact on graft and patient survival. Major postoperative complications were defined according to the Clavien-Dindo classification system

LIVER TRANSPLANTATION, Vol. 20, No. 8, 2014

DARIUS ET AL. 897

TABLE 4. Classification of BCs 1. Stricture

Anastomotic

Nonanastomotic

Isolated Diffuse

Mixed Ligated duct Anastomotic

2. Fistula

Nonanastomotic Missed duct

Partial Total Terminal Lateral

3. Other

Cholangitis Intrahepatic abscess Biliary lithiasis Aerobilia Hemobilia Secondary biliary cirrhosis

as grade 3 (requiring surgical, endoscopic, or radiological interventions) or 4 (life-threatening complications requiring intensive care treatment).

Statistical Methods Categorical variables were reported as numbers of cases and percentages. Pearson’s chi-square test and Fisher’s exact test were used for comparisons. Continuous variables were provided as medians and ranges. Their distributions were tested with the KolmogorovSmirnov test. The Student t test was adopted to compare the continuous variables with a parametric distribution. The Mann-Whitney U test was used for nonparametric variables. A P value < 0.05 indicated statistical significance. A univariate logistic regression analysis was performed with the intent of investigating the risk factors for the development of an anastomotic BC. Recipient risk factors included the following: recipient age, body surface, diagnosis of primary biliary atresia, and LT before 2002. Donor risk factors included sex and age. Technical risk factors included the following: type of liver graft (living versus deceased donor), Kasai procedure history, cold ischemia time, perfusion solution, graft RHAP, hepatic artery ligation during LT, number of arterial reconstructions and biliary anastomoses, and site of the biliary anastomosis. Follow-up risk factors included hepatic artery thrombosis, acute rejection, and CMV infections. Therefore, variables with a P value < 0.20 in the univariate analysis were used for the construction of a multivariate logistic regression model. The goodness of fit in the model was tested with the Hosmer-Lemeshow test. The risk predictions were

Stricture of the anastomosis identified by ultrasound as dilation of the bile ducts and often confirmed by a direct method (percutaneous or intraoperative cholangiography) Intrahepatic or extrahepatic (nonanastomotic) localized stricture Intrahepatic and extrahepatic (nonanastomotic) diffuse strictures Anastomotic and nonanastomotic strictures Ligated duct(s) over the graft’s cut surface Anastomotic fistula (50%) Peripheral, not from the graft’s cut surface From the cut surface, not connected to the main intrahepatic bile duct From the cut surface, connected to the main intrahepatic bile duct Clinically diagnosed without radiological findings of other BCs Radiological diagnosis, not other BCs

Not related to other causes of BCs

reported as P values, odds ratios (ORs), and 95% confidence intervals (CIs). Survival curves were produced with the Kaplan-Meier method. The log-rank test was used for comparisons between groups. Statistical analyses and plots were accomplished with SPSS 19.0 statistical software (SPSS, Chicago, IL).

RESULTS Overall Results The overall 1-, 5-, and 10-year patient and graft survival rates for the 429 pediatric LT recipients were 98%, 95%, and 94% and 97%, 94%, and 92%, respectively. The 1- and 5-year patient survival rates were 94% and 90%, 89% and 85%, 94% and 89%, and 98% and 94% for whole, reduced size, split, and living donor liver grafts, respectively (P 5 0.14). The 1- and 5-year graft survival rates were 89% and 87%, 79% and 76%, 83% and 78%, and 95% and 91% for whole, reduced size, split, and living donor liver grafts, respectively (P < 0.001).

BCs Table 5 summarizes the incidence of BCs according to our classification and the type of graft. The overall incidence of BCs was 23% (n 5 98); 61% of these BCs (n 5 60) were anastomotic, whereas 39% (n 5 38) were nonanastomotic. Anastomotic biliary strictures and fistulas accounted for 78% (n 5 47) and 22% (n 5 13) of the anastomotic BCs, respectively. Forty-two of the 98 BCs occurred within 3 months of LT (median 5 12 days, range 5 1-90 days), and 56 occurred more than 3 months after transplantation (median 5 257 days,

1

1 0 1 0

1 3

1 (1) 0 1 (1) 1 (1) 0

1 (1) 3 (3) 0

0 0

0

1 (1) 0

1

3 (3) 0 0 0

2

6 (7) 0 0 0

4

0 0

2 (2) 0 0

0

1 (1)

0 0 0

8 (9)

[n (%)]

Incidence

0 0

0 0

0

0

0

6

0

Treated (n)

6 (7)

Treated (n)

[n (%)]

Medically

2

2

1

8

Treated (n)

Surgically

1

0

0

0

Treated (n)

Medically

Reduced Size Liver (n 5 91)

*One was treated percutaneously, and the other was treated endoscopically.

Stricture Anastomotic Nonanastomotic Isolated Intrahepatic Extrahepatic Diffuse (intrahepatic 1 extrahepatic) Mixed (anastomotic and nonanastomotic) Ligated duct Fistula Anastomotic Partial (50%) Nonanastomotic Missed duct Terminal Lateral Other Cholangitis Intrahepatic abscess Hemobilia Secondary biliary cirrhosis Aerobilia Lithiasis

Surgically

Incidence

Whole Liver (n 5 88)

0 0

1 (2) 0 0 0

0 0

2 (4) 0 0

0

0

0 0 1 (2)

4 (9)

[n (%)]

Incidence

0

2

1

1

0

0

2 (1) 3 (2)

3 (2) 1 (1) 0 0

5 (3) 3 (2)

5 (3) 0 0

1 (1)

1 (1)

3 (2) 0 0

29 (14)

(n)

(n)

1

[n (%)]

Treated

Treated

3

Incidence

Medically

Surgically

Split Liver (n 5 47)

0 1

1 0

5 3

5

1

1

1

29

Treated (n)

Surgically

2 2*

2 1

0 0

0

0

2

0

Treated (n)

Medically

Living Donor (n 5 203)

TABLE 5. Incidence and Treatment of BCs According to the Proposed Classification and Graft Type for a Series of 429 Pediatric LT Procedures Performed at Saint-Luc University Clinic: July 1993 to November 2010

898 DARIUS ET AL. LIVER TRANSPLANTATION, August 2014

LIVER TRANSPLANTATION, Vol. 20, No. 8, 2014

DARIUS ET AL. 899

TABLE 6. Univariate Logistic Regression Analysis of the Risk of Anastomotic BCs for a Series of 429 Pediatric LT Procedures Performed at Saint-Luc University Clinic: July 1993 to November 2010 Variable Recipient features Age < 2 years Body surface (m2)  0.5 Primary biliary atresia LT before 2002 Donor features Sex: male Age < 2 years Age > 50 years Technical aspects Living donation Split liver Reduced liver Whole liver Kasai procedure before transplantation Cold ischemia time > 6 hours HTK perfusion liquid Retrograde arterial perfusion Arterial ligation at LT >1 arterial reconstruction >1 biliary anastomosis Duct-to-duct anastomosis Complications Hepatic artery thrombosis Acute rejection CMV infection

range 5 96-3480 days). The spectrum of early biliary strictures and fistulas was as follows: anastomotic strictures (n 5 10), intrahepatic isolated nonanastomotic strictures (n 5 1), mixed strictures (n 5 2), partial anastomotic fistulas (n 5 9), and total anastomotic fistulas (n 5 3). The spectrum of late biliary strictures and fistulas was as follows: anastomotic strictures (n 5 37), intrahepatic isolated nonanastomotic strictures (n 5 3), diffuse nonanastomotic strictures (n 5 2), mixed strictures (n 5 1), and partial anastomotic fistulas (n 5 1). No perforations or stenoses of the Roux-en-Y loop of the biliary reconstruction were observed. According to the univariate logistic regression analysis, only hepatic artery thrombosis and acute rejection were significant risk factors for anastomotic BCs (Table 6). The type of graft was not found to be a risk factor for the development of BCs. According to the multivariate logistic regression analysis, 3 independent risk factors for anastomotic complications were observed: hepatic artery thrombosis, acute rejection, and living donation (Table 7). Separately analyzing the risk factors for anastomotic strictures, we found that hepatic artery thrombosis (OR 5 1.03, P 5 0.02), acute rejection (OR 5 2.73, P 5 0.07), and living donation (OR 5 2.07, P 5 0.03) were statistically significant. Hepatic artery thrombosis (OR 5 9.94, P 5 0.001) was the only risk factor for anastomotic fistulas.

P Value

OR

95% CI

0.60 0.83 0.93 0.33

1.16 0.94 1.02 1.32

0.67-2.02 0.53-1.68 0.59-1.78 0.76-2.29

0.84 0.35 0.80

0.94 0.50 0.76

0.55-1.63 0.11-2.15 0.09-6.23

0.12 0.80 0.35 0.43 0.40 0.35 0.74 0.59 0.77 0.97 0.98 0.42

1.55 0.89 0.71 0.75 1.26 0.77 1.24 0.82 0.83 1.03 1.01 0.54

0.89-2.68 0.36-2.19 0.34-1.46 0.36-1.54 0.73-2.18 0.44-1.33 0.35-4.43 0.41-1.66 0.24-2.86 0.29-3.60 0.47-2.16 0.12-2.37

0.003 0.006 0.19

3.65 2.31 1.53

1.55-8.63 1.27-4.19 0.80-2.91

Impact of RHAP A subanalysis comparing the outcomes of living donor LT recipients with atraumatic RHAP (n 5 63) and without atraumatic RHAP (n 5 140) was performed. The incidence of BCs was 29% (n 5 41) for the 140 cases without RHAP and 24% (n 5 15) for the 63 cases with RHAP (P 5 0.570). At the 2-year follow-up, the rates of anastomotic biliary strictures were 6% (4/63) and 16% (22/ 140) with and without the use of RHAP, respectively (P 5 0.11). Hepatic artery thrombosis was observed only in the group without RHAP (3% versus 0%, P 5 0.63).

Management of Anastomotic BCs All anastomotic BCs except one were primarily surgically treated. A biliary drain was left in place during the surgical redo procedure for only 5 of the 59 patients (according to the surgeon’s decision during the operation). This biliary drain was placed transanastomotically, exteriorized through the Roux-en-Y loop, and fixed to the abdominal wall. The biliary drain was successfully removed after 3 months. Five of the 59 surgically treated patients (8.5%) developed major complications related to the reintervention: 4 required at least 1 extra surgical intervention because of biliary peritonitis (n 5 2), anastomotic fistulas (n 5 1), or a subobstruction (n 5 1; this patient needed

900 DARIUS ET AL.

LIVER TRANSPLANTATION, August 2014

TABLE 7. Multivariate Logistic Regression Analysis for the Risk of Anastomotic BCs for a Series of 429 Pediatric LT Procedures Performed at Saint-Luc University Clinic: July 1993 to November 2010 Variable

P Value

OR

95% CI

Hepatic artery thrombosis Acute rejection Living donation

Risk factors and surgical management of anastomotic biliary complications after pediatric liver transplantation.

Biliary complications (BCs) still remain the Achilles heel of liver transplantation (LT) with an overall incidence of 10% to 35% in pediatric series. ...
197KB Sizes 0 Downloads 3 Views