Iatrogenic bile duct injuries (IBDI) are a potentially severe and complex surgical complication that commonly occurs during laparoscopic cholecystectomy or gastroduodenal surgery.1 IBDI is defined as unintentional damage to any part of the biliary tree during a surgical intervention, which is frequently associated with leaks or obstruction.2
IBDI occurs most commonly during cholecystectomy, with incidence rates varying from 0.08% to 1.5%.3-6 The majority of these complications are initially treated with Roux-en-Y hepaticojejunostomy with success rates approaching 90%.2 Reconstruction after IBDI can progress to serious complications, including strictures, liver atrophy, recurrent cholangitis, intrahepatic lithiasis, and secondary biliary cirrhosis (SBC) with or without portal hypertension, resulting in the need for liver transplantation (LT).1,2 Furthermore, biliary injuries associated with vascular injuries may result in hepatic necrosis, intrahepatic abscess formation, and fulminant hepatic failure, accelerating the need for LT.7 The incidence of SBC after IBDI ranges from 7% to 25% in different series.2 There are only a limited number of case series that report on the utilization of LT in patients with complications of biliary reconstruction after IBDI.3-5,7
The indications, timing, and outcomes of LT in these patients require further investigation. The aim of this study is to evaluate perioperative and long-term outcomes of patients undergoing LT for IBDI in the United States using the United Network for Organ Sharing (UNOS) data set.
MATERIALS AND METHODS
The UNOS database was queried for all liver transplants performed in the United States between October 1994 and January 2014. Of the 101 238 liver transplants performed during the study period, 61 were for complications of biliary injury. We used the “Primary Diagnosis/Text of Other, Specify at Listing” file of the UNOS data set with the following keywords to identify the cases: Bile duct injury or damage, transected bile duct, bile duct injury after cholecystectomy, choledocal injury and complications after cholecystectomy. For patients identified with IBDI, collected data included the following: recipient age, sex, serum creatinine, total bilirubin, International Normalized Ratio (INR), Model End-Stage Liver Disease (MELD) score, coexisting liver disease, body mass index (BMI), blood type, pretransplant portal vein thrombosis, length of hospital stay, number of living donors, waiting time, acute rejection at 6 months, early graft loss, and overall graft and patient survival. BMI was classified as 18.5 or less, 18.6 to 25, 25.1 to 30, 30.1 to 40, and greater than 40 kg/mt2; cold ischemia time (CIT) as 6 or less, 6.1 to 12, and longer than 12 hours; and MELD as 6.1 to 10, 10.1 to 20, 20.1 to 30, 30.1 to 40, and greater than 40.
Donor characteristics, such as age, sex, and CIT were also collected. Pediatric transplants were excluded from the analysis. We performed a subgroup analysis comparing outcomes of patients transplanted with SBC after IBDI with patients transplanted for benign liver diseases, such as nonalcoholic steatohepatitis (NASH), primary sclerosing cholangitis (PSC), primary biliary cholangitis (PBC), and alcoholic liver disease. Our main end points were early graft loss defined as graft survival less than 7 days, perioperative mortality (30 days), graft and patient survival, and length of hospital stay.
Univariate analyses were performed using chi-squared or Fisher exact test as appropriate for categorical variables. Nonparametric continuous variables were reported by means (range) and compared using the Mann-Whitney U test. We performed a case-matched analysis to compare IBDI to non-IBDI patients adjusting for recipient age, sex, hepatitis C virus (HCV), MELD, hepatocellular carcinoma (HCC), CIT, and transplant era in a 5:1 ratio.
Survival was calculated from the time of transplantation using the Kaplan-Meier method with the log rank test to analyze differences in survival. Survival was also assessed after case-matched analysis. Logistic regression was performed to find independent predictors of early graft loss and perioperative mortality in both the overall cohort and using case matched data. A Cox regression analysis was performed to identify predictors of graft and patient survival after transplantation for IBDI. Factors found to be significant in the bivariate analysis were included in the multivariable analysis. Missing data of continuous variables were inputted with the mean using SPSS automatic imputation method. For most variables (creatinine, total bilirubin, LOS, BMI, and so on), the percentage of missing data was less than 2%. MELD score was missing in around 20% of the overall LT cases due to inadequate reporting of INR before the MELD era. Candidate models were chosen using stepwise forward selection and backward elimination to isolate the impact of IBDI. The results of the Cox models were presented as adjusted hazard ratios (HR) accompanied by 95% confidence interval (CI). Results from the logistic regression were presented as odds ratios accompanied by 95% CI. All tests were 2-sided with significance defined as P value less than 0.05. Analyses were conducted using SPSS software version 22.0.6 (SPSS Inc., Chicago, IL).
Patient Characteristics and Demographics
During the study period, 61 (0.06%) patients were transplanted for IBDI. In the IBDI group, 28 (45.9%) were men with a mean age of 50.16 ± 11.7 years. The median follow-up was 4.8 years in the IBDI group. The mean length of hospital stay was 19.1 days. The mean MELD score at the time of transplantation for this group was 22.6 ± 9.8 with a mean wait time of 224.6 ± 395 days. In the IBDI group, 12 (12.7%) patients were transplanted with MELD score exception. The mean creatinine, total bilirubin, and INR was 1.4 mg/dL, 13.9 mg/dL and 1.8, respectively. The mean BMI was 25.5 kg/m2. Acute cellular rejection at 6 months was 14.8% for those transplanted for IBDI. The incidence of portal vein thrombosis in this group is 8.2%, and 3.3% received a transplant from a living donor. None of the patients with IBDI were found to have alcoholic liver disease, NASH, or HCC.
We compared patients undergoing LT for IBDI with patients transplanted for other indications. Patients receiving LT after IBDI were more likely to be women (IBDI group, 54.09% vs other indications group; 34.48%, P = 0.001) and had less incidence of hepatitis C virus infection (IBDI 4.9% vs other 39.4%, P = 0.001). Total cold ischemic time was significantly higher in the IBDI group (P < 0.001). Creatinine levels were significantly lower, but total bilirubin was higher in patients with IBDI (P <0.001 for both). The rest of the demographics and patient characteristics are summarized in Table 1.
Perioperative mortality (30 days) was significantly higher in patients with LT after IBDI (13.1%) than in patients transplanted for other indications (4.3%) (P = 0.004). The 90-day mortality rate was 13.1% for patients transplanted with IBDI versus 6.9% (P = 0.071) in those transplanted for other indications. In Figure 1B, we can identify a pattern of decreased perioperative mortality in the IBDI group compared with patients transplanted for other indications.
We performed a multivariate analysis to find predictors of overall graft and patient survival in patients undergoing LT for IBDI. Men were found to be significantly associated with graft and patient survival in this cohort (P = 0.002; 95% CI, 1.52-6.71; HR, 3.19 and P = 0.016; HR, 3.57; 95% CI, 1.27-10.02). Patients with NASH, cholestatic liver disease (PSC, PBC), and alcoholic liver disease have similar outcomes to those transplanted for IBDI.
The most common causes of early graft failure in the IBDI group were primary graft nonfunction (n = 4), followed by multisystem organ failure (n = 1), portal vein stricture with biliary stricture (n = 1), vascular thrombosis (n = 1), and graft failure unspecified (n = 4). The retransplantation rate was 17.7%. The most common causes of perioperative death in the IBDI group were sepsis with multiorgan failure (n = 4), myocardial infarction (n = 1), ischemic bowel (n = 1), primary graft nonfunction (n = 1), and bleeding (n = 1).
Most of these cases with IBDI were transplanted in the UNOS region 3 (19.7%) conformed by Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, and Puerto Rico and region 5 (19.7%), conformed by California, Arizona, Nevada, New Mexico, and Utah. Other regions with a high percentage of patients with IBDI were region 7 (11.5%), conformed by Illinois, Minnesota, North Dakota, South Dakota, and Wisconsin; region 11 (11.5%) conformed by Kentucky, North Carolina, South Carolina, Tennessee and Virginia; and region 10 (9.8%), conformed by Indiana, Michigan, and Ohio.
Multivariable Regression Analysis of Prognostic Factors Associated With Early Graft Loss (<7 days) and Perioperative Mortality After LT in the United States
A multivariate logistic regression analysis was performed to find factors associated with early graft loss in patients who underwent LT in the United States. IBDI was identified as a very strong predictor of early graft loss (P <0.001; OR, 5.120 CI 3.08-8.79). Cold ischemic time, recipient BMI and MELD score were also significantly associated with increased risk of early graft loss after LT. This is reflected in Figure 1A where graft survival in the IBDI group is decreased, especially in the perioperative time.
A multivariate analysis was performed to find factors associated with perioperative mortality at 30 and 90 days in patients after LT. IBDI was identified as an independent factor associated with increased 30-day mortality (P = 0.005; 95% CI, 1.40-6.27; OR, 2.96) (Table 3). Other significant factors associated with increased risk of perioperative mortality were cold ischemic time and mean MELD score. Table 3 summarized 7-, 30-day, 1-, 3-, and 5-year graft and patient survivals after transplantation.
We performed a case-matched analysis comparing patients with and without IBDI in a 5:1 ratio adjusting for recipient age, sex, HCV, MELD, HCC, CIT, and transplant era (Table 1). The multivariable analysis after adjustment demonstrated similar results to those in the overall data. IBDI significantly increased the risk of early graft failure (OR, 8.4; P < 0.001) and 30-day perioperative mortality (OR, 2.9; P = 0.031). Other important risk factors included cold ischemia time for both early graft failure and perioperative mortality and MELD score for perioperative mortality (Table 2). We found no significant difference in graft and patient survival in the long term (Figures 2 A and B). Graft and patient survival rates after case matching were similar to those found in the overall cohort at 7-, 30-day, 1-, 3-, and 5-year cutoffs.
The incidence of biliary complication after cholecystectomy has remained stable during the past decade. IBDI can potentially be very complex requiring multiple endoscopic, surgical, and percutaneous interventions.8-10 These injuries are associated with high postoperative morbidity and mortality after the initial biliary reconstruction.7 More recently, IBDIs tend to be more complicated involving the proximal biliary tree and not infrequently compromising vascular structures in the hepatic hilum sometimes requiring major surgical procedures, such as hepatectomy, for definitive treatment.8,10,11 We found that CIT was significantly longer in the IBDI group which maybe a surrogate of surgical complexity (prolonged hepatectomy).
Appropriate management of IBDI involves early recognition; correct classification, to assess the extent and type of injury; advanced surgical technique with appropriate repair by a hepatobiliary surgeon; and a multidisciplinary approach in a specialized center.12
LT is a rare indication for the treatment of complications arising from IBDI and its associated repair.1,8,13 Addeo et al8 reviewed a total of 20 studies from multiple institutions with a total of 53 patients who underwent LT after complications from IBDI. Most of the patients presented with irreversible complications after vascular injuries and previous failed repair attempts. Concomitant vascular injuries at the time of the IBDI have been reported as a significant prognostic factor, increasing the rates of perioperative morbidity and mortality after reconstruction.14
It has been reported that approximately 0.2% of patients will require LT after biliary complications after repair of IBDI.2,15 In a national survey of 18 centers in Argentina, Ardiles et al15 evaluated 16 patients with IBDI that required LT showing a perioperative mortality of 22%. Using the UNOS data set, we were able to identify 61 cases transplanted for IBDI, representing 0.06% of all cases transplanted in the United States during the study period.
Barbier et al2 found higher perioperative morbidity and mortality rates in patients transplanted for IBDI than those transplanted for other indications. Our risk-adjusted analysis demonstrated that patients undergoing transplantation for IBDI have eightfold increased rates of early graft loss (within 7 days) compared with patients transplanted for other indications. We speculate that the very significant increased rate of early graft loss in this cohort is at least in part the result of previous vascular injuries requiring more complex arterial and venous reconstructions during LT. Notably, patients undergoing LT for IBDI have similar 1-, 3-, and 5-year survival rates (84.3%, 84.3%, and 76.2%, respectively) than those transplanted for other indications (87.4%, 79.3%, and 72.9%, respectively). In addition, we found no significant differences in survival between patients transplanted in this cohort and those transplanted for benign diseases, including NASH, cholestatic liver disease (PSC and PBC), and alcoholic liver disease.
We reported a 30-day perioperative mortality of 13.1% in this cohort compared with 4.3% in patients transplanted for other indications. Our results pointed out that IBDI, as an indication for LT, is a strong independent predictor of perioperative mortality increasing almost threefold the risk of perioperative death. These results were similar after case-matched analysis after adjusting for factors, such as age, sex, HCV, MELD, HCC, CIT, and transplant era. Sex was the only independent factor associated with decreased overall graft survival in patients undergoing LT for IBDI. Notably, the majority of these cases were transplanted in the centers from UNOS regions 3, 5, 7, 10, and 11. This could be related to the high number of referral centers in those areas. Sepsis and multiorgan failure were the most common causes of perioperative mortality in this cohort which may be related to multiple percutaneous interventions, bacterial colonization, and prolonged operative time and complexity.
There are several limitations that should be discussed. Variables, such as time of the biliary injury to initial repair, type of surgical reconstruction, and preoperative percutaneous procedures performed, are not included in the UNOS database. Type and classification along with the coexistence of vascular injury, rates of recurrent cholangitis, bacteriology, preoperative antibiotic treatment, and type of perioperative morbidity, that could help interpret the causes of perioperative deaths are not captured in this data set. The strength of our study not only comes from the significant number of patients transplanted for this indication included in a single database but also from its clinically robust and uniform definitions of patient characteristics and events.
In summary, LT in patients with IBDI is rare. Excellent long-term outcomes suggest that LT is a reasonable option in highly selected patients with IBDI who have acute liver failure or significant complications after biliary reconstruction. Those transplanted for IBDI are at a significantly increased risk of early graft loss with increased rates of retransplantation. Perioperative mortality is also significantly increased, demonstrating important perioperative challenges possibly associated with increased operative complexity in patients with major right upper quadrant surgery before LT. The long-term results of LT after IBDI are comparable to those patients transplanted for other benign indications, such as NASH, PSC, PBC, and acute liver disease. Future studies are needed to prospectively investigate causes of perioperative morbidity and to identify potential modifiable factors to improve outcomes in patients undergoing transplantation for IBDI.
1. Thomson BN, Parks RW, Madhavan KK, et al. Liver resection and transplantation in the management of iatrogenic biliary injury. World J Surg
2. Barbier L, Souche R, Slim K, et al. Long-term consequences of bile duct injury after cholecystectomy. J Visc Surg
3. Eikermann M, Siegel R, Broeders I, et al. Prevention and treatment of bile duct injuries during laparoscopic cholecystectomy: the clinical practice guidelines of the European Association for Endoscopic Surgery (EAES). Surg Endosc
4. Törnqvist B, Strömberg C, Persson G, et al. Effect of intended intraoperative cholangiography and early detection of bile duct injury on survival after cholecystectomy: population based cohort study. BMJ
5. Halbert C, Pagkratis S, Yang J, et al. Beyond the learning curve: incidence of bile duct injuries following laparoscopic cholecystectomy normalize to open in the modern era. Surg Endosc
6. Hariharan D, Psaltis E, Scholefield JH, et al. Quality of life and medico-legal implications following iatrogenic bile duct injuries. World J Surg
7. Robles Campos R, Marín Hernández C, Fernández Hernández JA, et al. [Delayed right hepatic artery haemorrhage after iatrogenic gallbladder by laparoscopic cholecystectomy that required a liver transplant due to acute liver failure: clinical case and review of the literature]. Cir Esp
8. Addeo P, Saouli AC, Ellero B, et al. Liver transplantation for iatrogenic bile duct injuries sustained during cholecystectomy. Hepatology international
9. Halbert C, Altieri MS, Yang J, et al. Long-term outcomes of patients with common bile duct injury following laparoscopic cholecystectomy. Surg Endosc
10. Hogan NM, Dorcaratto D, Hogan AM, et al. Iatrogenic common bile duct injuries: Increasing complexity in the laparoscopic era: a prospective cohort study. Int J Surg
. 2016;33 Pt A:151–156.
11. Jablonska B. Hepatectomy for bile duct injuries: when is it necessary? World J Gastroenterol
12. Lau WY, Lai EC, Lau SH. Management of bile duct injury after laparoscopic cholecystectomy: a review. ANZ J Surg
13. Lubikowski J, Chmurowicz T, Post M, et al. Liver transplantation as an ultimate step in the management of iatrogenic bile duct injury complicated by secondary biliary cirrhosis. Ann Transplant
14. Hafeez Bhatti AB, Dar FS, Zia H, et al. Prognostication of learning curve on surgical management of vasculobiliary injuries after cholecystectomy. Int J Hepatol
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
15. Ardiles V, McCormack L, Quiñonez E, et al. Experience using liver transplantation for the treatment of severe bile duct injuries over 20 years in Argentina: results from a National Survey. HPB (Oxford)