150
Original article
Predictors of endoscopic treatment outcome in the management of biliary complications after orthotopic liver transplantation Giacomo Faleschinia, Salvatore F. Vadalà di Pramperoa, Milutin Bulajica,d, Umberto Baccaranib, Pierluigi Toniuttoc, Nikola Panice, Loris M. Zorattia, Marco Marinoa and Maurizio Zillia Background and aims The most common complications after liver transplantation nowadays affect the biliary tract. We carried out a retrospective study to identify predictors of endoscopic treatment outcome in the management of post-transplantation biliary complications. Methods Data from all patients with post-transplantation biliary complications subjected to endoscopic treatment at the University of Udine between 2000 and 2012 were extracted. To identify predictors of endoscopic treatment outcome, a logistic regression analysis was carried out. Cox modeling was used to identify factors associated with mortality. Results We identified 142 patients who developed biliary complications: 83 of these patients had a successful endoscopic therapy, whereas 45 had a failure. Fourteen patients, who developed nonanastomotic biliary stricture, were excluded from the analysis. Patients with biliary complications who had pretransplant Model for End-Stage Liver Disease score more than 10 [odds ratio (OR) 3.88; 95% confidence interval (CI) 1.16–12.95; P = 0.03] and stent retention time more than 12 months (OR 6.45; 95% CI 2.14–19.42; P < 0.01) were less likely to respond to endoscopic therapy. In contrast, both dilatation and stenting procedures (OR 0.10; 95% CI 0.03–0.30; P < 0.01) and 10 Fr diameter stent placement (OR 0.21; 95% CI 0.07–0.70; P = 0.01) predicted favorable endoscopic treatment outcome. Time to the occurrence of biliary complications
Introduction Biliary complications are among the most common problems following liver transplantation (LT). Nowadays, they occur in up to 25% of cases, with a mortality of 10% following LT and 28–32% following LT from an adult living donor [1–3]. Post-transplantation biliary complications include anastomotic strictures (ASs), nonanastomotic biliary strictures (NASs), bile leaks, choledocholithiasis, cast syndrome, biloma, hemobilia, and sphincter of Oddi dysfunction. In the past, surgical revision was considered to be the gold standard in resolving post-transplantation biliary complications. However, endoscopic retrograde cholangiopancreatography (ERCP) has recently been 0954-691X © 2015 Wolters Kluwer Health | Lippincott Williams & Wilkins
of more than 3 months [hazard ratio (HR) 0.24; 95% CI 0.10–0.56] and placement of five or more stents (HR 0.31; 95% CI 0.12–0.79) were found to be protective against mortality, whereas hepatic artery thrombosis was a significant risk factor for mortality (HR 13.88; 95% CI 4.08–47.25). Conclusion We found endoscopic treatment to be less effective in patients with pretransplant Model for End-Stage Liver Disease score more than 10 and stent retention time more than 12, whereas dilatation and stenting procedure and 10 Fr diameter stent placement predicted a favorable outcome. Eur J Gastroenterol Hepatol 27:150–154 © 2015 Wolters Kluwer Health | Lippincott Williams & Wilkins. European Journal of Gastroenterology & Hepatology 2015, 27:150–154 Keywords: biliary tract, endoscopic retrograde cholangiopancreatography, stent, strictures, survival a Department of Gastroenterology and Digestive Endoscopy, Academic Center ‘Santa Maria della Misericordia’, bLiver Transplantation Unit, Department of Surgery, cMedical Liver Transplant Section, Department of Internal Medicine, University of Udine, Udine, dFaculty of Medicine, University of Belgrade and e University Clinical-Hospital Center ‘Dr Dragisa Misovic-Dedinje’, Belgrade, Serbia
Correspondence to Giacomo Faleschini, MD, Department of Gastroenterology and Digestive Endoscopy, Academic Center ‘Santa Maria della Misericordia’, Piazzale S. Maria della Misericordia, 15, 33100 Udine, Italy Tel: + 39 0432 552581/2583; fax: + 39 0432 552588; e-mail: [email protected] Received 17 September 2014 Accepted 31 October 2014
shown to be effective in the treatment of ASs, choledocholithiasis, and bile leaks [4,5]. Surgical factors, such as improper surgical techniques, a mismatch in size between the donor and the recipient bile ducts, inappropriate suture material, tension at the anastomosis, excessive use of electrocauterization for control of bleeding, infection, and ischemia, have been associated previously with the development of posttransplantation AS [6–8]. However, hepatic artery thrombosis, chronic ductopenic rejection, prolonged warm and cold ischemic time, as well as ABO blood group system incompatibility and donation after cardiac death have been linked to the occurrence of NAS [4]. Nevertheless, limited data are available so far on the DOI: 10.1097/MEG.0000000000000251
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Endoscopic outcome in liver transplantation Faleschini et al. 151
predictors of the outcome of the endoscopic treatment in patients with post-transplantation biliary complications. It has been shown previously that more aggressive approaches including the use of multiple stents and maximal balloon dilation can achieve more expeditious resolution of ASs, with a high success rate [9–11]. Furthermore, Buxbaum et al. [12] identified take-back surgery for a nonbiliary indication during the first month after LT, as well as transplantation from a donor after cardiac death or living donor to be associated with less favorable outcome of the endoscopic treatment. However, as the ERCP-guided endoscopic treatment is considered to be method of choice in the treatment of patients with post-transplantation biliary complications, more evidence is needed to identify factors leading to treatment success. The primary aim of this study was to identify predictors of ERCP-guided endoscopic treatment outcome in patients developing biliary complications after LT.
Patients and methods Study population and data extraction
To identify LT recipients with duct-to-duct anastomoses in need of therapeutic ERCP because of biliary complications, we retrospectively reviewed the medical records and endoscopic database of all adult patients who underwent LT at the University of Udine between 1 January 2000 and 31 December 2012. Only patients primarily subjected to the ERCP were eligible for inclusion. Patients who underwent percutaneous transhepatic cholangiography (PTC) or surgery first were not included in the study. The details of each transplantation surgery, including the cold and warm ischemic time and the use of a T-tube, were reviewed and recorded. Both donor age and recipient age were determined. Finally, the primary liver disease, laboratory pretransplant Model for EndStage Liver Disease (MELD) score was calculated retrospectively for each patient; the Child–Pugh score, graft steatosis, sex, pretransplant bilirubin, and detailed clinical course for all recipients were collected. Patients with NAS were not included either in the analysis of endoscopic treatment predictor factors or in the survival analysis. Transient strictures, which have been defined as postoperative strictures that did not require either endoscopic or PTC treatment, were excluded from the analysis. The study was carried out according to the Declaration of Helsinki. Endoscopic treatment
All patients underwent magnetic resonance cholangiopancreatography (MRCP) before the ERCP to assess the level and morphology of biliary strictures. ERCPs were performed by three expert endoscopists in an endoscopy unit that performs more than 500 procedures per year. All procedures were carried out under anesthesiologistassisted deep sedation (clinical monitoring of vital
functions, propofol, midazolam, fentanyl, N-hyoscine butylbromide i.v.). Endoscopic sphincterotomy (standard double-lumen sphincterotome Ultratome; Boston Scientific, Natick, Massachusetts, USA) was the main technique to access the bile duct. Precut sphincterotomy (Microknife XL Needleknife; Boston Scientific) was only performed when standard cannulation maneuvers failed to provide access. After selective cannulation of the common bile duct of the recipient, the stenosis was first passed by a hydrophilic guide wire (straight or angled type, 0.021–0.035′′, Radiofocus; Terumo, Somerset, New Jersey, USA), followed by contrastography of the occluded graft biliary tree. Patients with AS underwent a balloon dilation (4–10 mm in diameter; Boston Scientific) of the stricture, followed subsequently by the placement of the maximum number of insertable plastic stents, if possible (Amsterdam Type, 7–10 Fr; Boston Scientific). Repeat balloon dilation and stent exchange were performed every 2–3 months, until complete stricture resolution, which was confirmed by cholangiography using an inflated extraction balloon after stent removal. Biliary leaks were addressed by temporary placement of 7–10 Fr stents, which were exchanged if necessary after 2–3 months in case of leakage persistence. All patients received intravenous prophylactic antibiotics (amoxicillin 1 g i.v. or ciprofloxacin 750 mg oral single dose) at least 2 h before ERCP. Antibiotics were continued if the patients had evidence of previous infections or in case of incomplete biliary drainage. Follow-up consisted of clinical evaluation, liver function tests, MRCP 6 months after the final stents removal, and transabdominal ultrasonography every 3 months after the conclusion of the treatment during the first year and then every 6 months thereafter. Definition of findings and outcomes
Biliary complications were defined as an increase in liver function test results with definitive cholangiographic (endoscopic or percutaneous) evidence of an AS, leak, NAS, stones, or papillary stenosis. AS was defined as a dominant short narrowing at the anastomotic site without free or effective passage of contrast material, as shown by cholangiography. NAS was considered when one or more strictures occurred more than 0.5 cm proximal to the biliary anastomosis [4]. Bile leak was defined as the extravasation of contrast material into the peritoneum from the biliary anastomosis, the T-tube tract, or from the cut surface of the liver. Successful endoscopic therapy was defined as the normalization of liver function tests with no evidence of clinical symptoms of cholestasis and demonstration by cholangiography (MRCP), 6 months after stent removal, of stricture or leak resolution, not requiring further endoscopic stent placement for the management of biliary problems at least 1 year from the last procedure. Recurrence was defined as the reappearance of biliary
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
152 European Journal of Gastroenterology & Hepatology
2015, Vol 27 No 2
complications requiring additional procedures (including stent placement, PTC, or surgery) after previous stent removal. Statistical analysis
Continuous variables were summarized as means and SDs or medians and interquartile range depending on their apparent distributions. Categorical data were summarized as percentages. Potentially relevant risk factors were assessed by logistic regression. Factors associated with P value less than 0.05 in univariate analysis were evaluated by the multivariate logistic regression technique using stepwise forward selection. Associations are specified as odds ratio (OR) with a 95% confidence interval (CI). A two-sided P value of less than 0.05 was considered to be significant. Multivariate logistic regression analysis was then carried out again after excluding covariates that were outside the interval (− 2, 2) on Pearson’s and deviance residual analysis. Kaplan–Meier survival analysis was carried out censoring patients at the time of death or at the end of the study period to determine 1-, 3-, and 5-year patient survival. Multivariate analysis with Cox proportional-hazard modeling was used to calculate hazard ratios (HRs) adjusted for potential confounding factors. These included recipient sex, age, indication for transplant, pretransplant MELD score, warm and cold ischemia time, and donor age. Statistical analysis was carried out using STATA (version 12; StataCorp, College Station, Texas, USA).
Results Over the 13-year period, 360 adult patients underwent LT. Table 1 reports their demographic and clinical Demographic and clinical characteristics and technical features of the procedure of 360 patients subjected to liver transplantation
Table 1
N (%) Demographic and clinical characteristics Age (years)a Male Hepatitis B virus Hepatitis C virus Alcoholic cirrhosis Hepatocellular carcinoma Fulminant hepatitis Primary biliary cirrhosis Primary sclerosing cholangitis Cryptogenic cirrhosis Other causes Technical features of the transplantation Split liverb Total ischemia timea Repeat liver transplantation Renal liver transplantation Hepatic artery thrombosis Pretransplant bilirubina Pretransplant MELD scorea Graft steatosis > 30%
55 272 17 104 85 101 10 7 5 10 21
(48–60) (74.7) (4.7) (28.9) (23.6) (28.1) (2.8) (1.9) (1.4) (2.8) (5.8)
37 466 31 11 9 3.1 14.5 35
(10.2) (399–560) (8.5) (3.0) (2.5) (1.3–7.2) (9.0–22.1) (9.6)
MELD, Model for End-Stage Liver Disease. a Median and interquartile range. b All patients with split liver graft received a choledochocholedochostomy.
characteristics as well as transplantation technical features. All patients received a liver from a deceased donor. A total of 142 (39%) patients developed posttransplantation biliary complications and were subsequently primarily subjected to ERCP. Table 2 reports the clinical features and procedure characteristics of 142 patients who underwent ERCP. Among patients treated because of a single posttransplantation complication, the majority developed AS (n = 79, 55.6%), followed by leaks (n = 18, 12.7%) and NAS (n = 14, 9.9%). Nineteen (13.4%) patients were treated because of strictures combined with bile duct stones, whereas 12 (8.6%) developed strictures and leaks at the same time (Table 2). The 14 patients with NAS were treated by a percutaneous intervention after undergoing ERCP for diagnostic confirmation, whereas ERCP-guided endoscopic therapy was attempted in 128 (90.1%) patients. Five (3.9%) of these patients were treated using the rendezvous technique (data not shown). The fourteen patients, who developed NAS, were excluded from analysis. In total, 490 ERCPs were performed, with a median number of three ERCPs per patient (interquartile range 2–5) (data not shown). Patients with both stones and strictures had a median number of four ERCPs per patient (interquartile range 2–5), followed by 3.5 in those with leaks and strictures (interquartile range 2–7), whereas patients with only AS or leaks had a median of three procedures per patient (interquartile range 2–5 and 2–3.75, respectively) (Table 2). The maximal average of four stents per patients was placed in patients with both stones and strictures and they were changed until a median of 24 months (Table 2). Among 128 patients who underwent ERCP-guided endoscopic therapy, definitive success was achieved in 83 (65%) (data not shown). As for the specific subgroups, the best therapeutic response was found among patients with AS (68%) and worse among those with both leaks and strictures (50%) (Table 2). We recorded four (3.1%) complications in total. Three of those were in the form of postsphincterotomy bleeding in patients treated because of AS, whereas one involved perforation in a patient with post-transplantation bile leak (Table 2). All bleeding episodes responded to endoscopic treatment (two patients with a submucosal injection of adrenaline solution and one patient with hemoclip positioning) and the perforation case responded to conservative treatment, without the need for surgical management. Predictors of outcome of endoscopic retrograde cholangiopancreatography-guided endoscopic treatment
Table 3 reports the technical features of transplantation surgery and endoscopic procedure associated significantly with the outcome of ERCP-guided endoscopic treatment (Hosmer–Lemeshow’s goodness of fit P = 0.64). Patients with a pretransplant MELD score more than 10 (OR 3.88;
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Endoscopic outcome in liver transplantation Faleschini et al. 153
Clinical features and treatment characteristics of 142 patients who underwent endoscopic retrograde cholangiopancreatography because of post-transplantation biliary complications
Table 2
Type of complications
N
Number of procedures (median)
Number of stents (median) (months)a
Stents retention time (median) (months)
Endoscopic success (%)
Complications
Anastomotic stricture Nonanastomotic stricture Leak only Leak + stricture Stones + stricture
79 14 18 12 19
3 1 3 3.5 4
3 NA 2 3 4
10.5 NA 10 11.5 24
68 NA 61 50 63
3 (bleeding) NA 1 (perforation) 0 0
a
Refers to the total number of stents inserted during the entire treatment period.
Transplantation surgery and technical features of the endoscopic procedure significantly associated with failure of endoscopic retrograde cholangiopancreatography-guided endoscopic therapy in 128 patients treated because of post-transplantation biliary complications
Table 3
OR (95% CI)a Pretransplant MELD score > 10 Stents retention time > 12 months Procedure of dilatation and stenting Stents diameter of 10 Fr
3.88 6.45 0.10 0.21
(1.16–12.95) (2.14–19.42) (0.03–0.30) (0.07–0.70)
P value 0.03 < 0.01 < 0.01 0.01
CI, confidence interval; MELD, Model for End-Stage Liver Disease; OR, odds ratio. Adjusted for recipient sex and age, indication for transplant, type of biliary complication, total ischemia time, Model for End-Stage Liver Disease. a
95% CI 1.16–12.95), as well as those who had a stent retention time of more than 12 months (OR 6.45; 95% CI 2.14–19.42), were less likely to respond favorably to endoscopic therapy. However, patients subjected to both the stenting and dilatation procedures (OR 0.10; 95% CI 0.03–0.30), as well as those treated with 10 Fr stents were more likely to achieve resolution of the complications (OR 0.21; 95% CI 0.07–0.70). On univariate analysis, hepatitis C virus infection appeared not to be associated with endoscopic outcome (P = 0.70).
Survival analyses
Of the 128 patients who were treated by ERCP in this study, none was lost to follow-up. Overall estimated 1-, 3-, and 5-year patient survival was 92, 82, and 77%, respectively (data not shown). In December 2012, 93 patients were still living, with a median follow-up from LT of 5.1 years (interquartile range 2.1–8.1 years) and 4.2 years (interquartile range 1.9–6.9 years) from first ERCP. Table 4 Factors associated significantly with survival and mortality in 128 patients treated with endoscopic retrograde cholangiopancreatography-guided endoscopic therapy because of post-transplantation biliary complications
Time to complication > 3 months Number of stent placed ≥ 5 Hepatic artery thrombosis
HR (95% CI)a
P value
0.24 (0.10–0.56) 0.31 (0.12–0.79) 13.88 (4.08–47.25)
< 0.01 0.02
Original article
Predictors of endoscopic treatment outcome in the management of biliary complications after orthotopic liver transplantation Giacomo Faleschinia, Salvatore F. Vadalà di Pramperoa, Milutin Bulajica,d, Umberto Baccaranib, Pierluigi Toniuttoc, Nikola Panice, Loris M. Zorattia, Marco Marinoa and Maurizio Zillia Background and aims The most common complications after liver transplantation nowadays affect the biliary tract. We carried out a retrospective study to identify predictors of endoscopic treatment outcome in the management of post-transplantation biliary complications. Methods Data from all patients with post-transplantation biliary complications subjected to endoscopic treatment at the University of Udine between 2000 and 2012 were extracted. To identify predictors of endoscopic treatment outcome, a logistic regression analysis was carried out. Cox modeling was used to identify factors associated with mortality. Results We identified 142 patients who developed biliary complications: 83 of these patients had a successful endoscopic therapy, whereas 45 had a failure. Fourteen patients, who developed nonanastomotic biliary stricture, were excluded from the analysis. Patients with biliary complications who had pretransplant Model for End-Stage Liver Disease score more than 10 [odds ratio (OR) 3.88; 95% confidence interval (CI) 1.16–12.95; P = 0.03] and stent retention time more than 12 months (OR 6.45; 95% CI 2.14–19.42; P < 0.01) were less likely to respond to endoscopic therapy. In contrast, both dilatation and stenting procedures (OR 0.10; 95% CI 0.03–0.30; P < 0.01) and 10 Fr diameter stent placement (OR 0.21; 95% CI 0.07–0.70; P = 0.01) predicted favorable endoscopic treatment outcome. Time to the occurrence of biliary complications
Introduction Biliary complications are among the most common problems following liver transplantation (LT). Nowadays, they occur in up to 25% of cases, with a mortality of 10% following LT and 28–32% following LT from an adult living donor [1–3]. Post-transplantation biliary complications include anastomotic strictures (ASs), nonanastomotic biliary strictures (NASs), bile leaks, choledocholithiasis, cast syndrome, biloma, hemobilia, and sphincter of Oddi dysfunction. In the past, surgical revision was considered to be the gold standard in resolving post-transplantation biliary complications. However, endoscopic retrograde cholangiopancreatography (ERCP) has recently been 0954-691X © 2015 Wolters Kluwer Health | Lippincott Williams & Wilkins
of more than 3 months [hazard ratio (HR) 0.24; 95% CI 0.10–0.56] and placement of five or more stents (HR 0.31; 95% CI 0.12–0.79) were found to be protective against mortality, whereas hepatic artery thrombosis was a significant risk factor for mortality (HR 13.88; 95% CI 4.08–47.25). Conclusion We found endoscopic treatment to be less effective in patients with pretransplant Model for End-Stage Liver Disease score more than 10 and stent retention time more than 12, whereas dilatation and stenting procedure and 10 Fr diameter stent placement predicted a favorable outcome. Eur J Gastroenterol Hepatol 27:150–154 © 2015 Wolters Kluwer Health | Lippincott Williams & Wilkins. European Journal of Gastroenterology & Hepatology 2015, 27:150–154 Keywords: biliary tract, endoscopic retrograde cholangiopancreatography, stent, strictures, survival a Department of Gastroenterology and Digestive Endoscopy, Academic Center ‘Santa Maria della Misericordia’, bLiver Transplantation Unit, Department of Surgery, cMedical Liver Transplant Section, Department of Internal Medicine, University of Udine, Udine, dFaculty of Medicine, University of Belgrade and e University Clinical-Hospital Center ‘Dr Dragisa Misovic-Dedinje’, Belgrade, Serbia
Correspondence to Giacomo Faleschini, MD, Department of Gastroenterology and Digestive Endoscopy, Academic Center ‘Santa Maria della Misericordia’, Piazzale S. Maria della Misericordia, 15, 33100 Udine, Italy Tel: + 39 0432 552581/2583; fax: + 39 0432 552588; e-mail: [email protected] Received 17 September 2014 Accepted 31 October 2014
shown to be effective in the treatment of ASs, choledocholithiasis, and bile leaks [4,5]. Surgical factors, such as improper surgical techniques, a mismatch in size between the donor and the recipient bile ducts, inappropriate suture material, tension at the anastomosis, excessive use of electrocauterization for control of bleeding, infection, and ischemia, have been associated previously with the development of posttransplantation AS [6–8]. However, hepatic artery thrombosis, chronic ductopenic rejection, prolonged warm and cold ischemic time, as well as ABO blood group system incompatibility and donation after cardiac death have been linked to the occurrence of NAS [4]. Nevertheless, limited data are available so far on the DOI: 10.1097/MEG.0000000000000251
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Endoscopic outcome in liver transplantation Faleschini et al. 151
predictors of the outcome of the endoscopic treatment in patients with post-transplantation biliary complications. It has been shown previously that more aggressive approaches including the use of multiple stents and maximal balloon dilation can achieve more expeditious resolution of ASs, with a high success rate [9–11]. Furthermore, Buxbaum et al. [12] identified take-back surgery for a nonbiliary indication during the first month after LT, as well as transplantation from a donor after cardiac death or living donor to be associated with less favorable outcome of the endoscopic treatment. However, as the ERCP-guided endoscopic treatment is considered to be method of choice in the treatment of patients with post-transplantation biliary complications, more evidence is needed to identify factors leading to treatment success. The primary aim of this study was to identify predictors of ERCP-guided endoscopic treatment outcome in patients developing biliary complications after LT.
Patients and methods Study population and data extraction
To identify LT recipients with duct-to-duct anastomoses in need of therapeutic ERCP because of biliary complications, we retrospectively reviewed the medical records and endoscopic database of all adult patients who underwent LT at the University of Udine between 1 January 2000 and 31 December 2012. Only patients primarily subjected to the ERCP were eligible for inclusion. Patients who underwent percutaneous transhepatic cholangiography (PTC) or surgery first were not included in the study. The details of each transplantation surgery, including the cold and warm ischemic time and the use of a T-tube, were reviewed and recorded. Both donor age and recipient age were determined. Finally, the primary liver disease, laboratory pretransplant Model for EndStage Liver Disease (MELD) score was calculated retrospectively for each patient; the Child–Pugh score, graft steatosis, sex, pretransplant bilirubin, and detailed clinical course for all recipients were collected. Patients with NAS were not included either in the analysis of endoscopic treatment predictor factors or in the survival analysis. Transient strictures, which have been defined as postoperative strictures that did not require either endoscopic or PTC treatment, were excluded from the analysis. The study was carried out according to the Declaration of Helsinki. Endoscopic treatment
All patients underwent magnetic resonance cholangiopancreatography (MRCP) before the ERCP to assess the level and morphology of biliary strictures. ERCPs were performed by three expert endoscopists in an endoscopy unit that performs more than 500 procedures per year. All procedures were carried out under anesthesiologistassisted deep sedation (clinical monitoring of vital
functions, propofol, midazolam, fentanyl, N-hyoscine butylbromide i.v.). Endoscopic sphincterotomy (standard double-lumen sphincterotome Ultratome; Boston Scientific, Natick, Massachusetts, USA) was the main technique to access the bile duct. Precut sphincterotomy (Microknife XL Needleknife; Boston Scientific) was only performed when standard cannulation maneuvers failed to provide access. After selective cannulation of the common bile duct of the recipient, the stenosis was first passed by a hydrophilic guide wire (straight or angled type, 0.021–0.035′′, Radiofocus; Terumo, Somerset, New Jersey, USA), followed by contrastography of the occluded graft biliary tree. Patients with AS underwent a balloon dilation (4–10 mm in diameter; Boston Scientific) of the stricture, followed subsequently by the placement of the maximum number of insertable plastic stents, if possible (Amsterdam Type, 7–10 Fr; Boston Scientific). Repeat balloon dilation and stent exchange were performed every 2–3 months, until complete stricture resolution, which was confirmed by cholangiography using an inflated extraction balloon after stent removal. Biliary leaks were addressed by temporary placement of 7–10 Fr stents, which were exchanged if necessary after 2–3 months in case of leakage persistence. All patients received intravenous prophylactic antibiotics (amoxicillin 1 g i.v. or ciprofloxacin 750 mg oral single dose) at least 2 h before ERCP. Antibiotics were continued if the patients had evidence of previous infections or in case of incomplete biliary drainage. Follow-up consisted of clinical evaluation, liver function tests, MRCP 6 months after the final stents removal, and transabdominal ultrasonography every 3 months after the conclusion of the treatment during the first year and then every 6 months thereafter. Definition of findings and outcomes
Biliary complications were defined as an increase in liver function test results with definitive cholangiographic (endoscopic or percutaneous) evidence of an AS, leak, NAS, stones, or papillary stenosis. AS was defined as a dominant short narrowing at the anastomotic site without free or effective passage of contrast material, as shown by cholangiography. NAS was considered when one or more strictures occurred more than 0.5 cm proximal to the biliary anastomosis [4]. Bile leak was defined as the extravasation of contrast material into the peritoneum from the biliary anastomosis, the T-tube tract, or from the cut surface of the liver. Successful endoscopic therapy was defined as the normalization of liver function tests with no evidence of clinical symptoms of cholestasis and demonstration by cholangiography (MRCP), 6 months after stent removal, of stricture or leak resolution, not requiring further endoscopic stent placement for the management of biliary problems at least 1 year from the last procedure. Recurrence was defined as the reappearance of biliary
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
152 European Journal of Gastroenterology & Hepatology
2015, Vol 27 No 2
complications requiring additional procedures (including stent placement, PTC, or surgery) after previous stent removal. Statistical analysis
Continuous variables were summarized as means and SDs or medians and interquartile range depending on their apparent distributions. Categorical data were summarized as percentages. Potentially relevant risk factors were assessed by logistic regression. Factors associated with P value less than 0.05 in univariate analysis were evaluated by the multivariate logistic regression technique using stepwise forward selection. Associations are specified as odds ratio (OR) with a 95% confidence interval (CI). A two-sided P value of less than 0.05 was considered to be significant. Multivariate logistic regression analysis was then carried out again after excluding covariates that were outside the interval (− 2, 2) on Pearson’s and deviance residual analysis. Kaplan–Meier survival analysis was carried out censoring patients at the time of death or at the end of the study period to determine 1-, 3-, and 5-year patient survival. Multivariate analysis with Cox proportional-hazard modeling was used to calculate hazard ratios (HRs) adjusted for potential confounding factors. These included recipient sex, age, indication for transplant, pretransplant MELD score, warm and cold ischemia time, and donor age. Statistical analysis was carried out using STATA (version 12; StataCorp, College Station, Texas, USA).
Results Over the 13-year period, 360 adult patients underwent LT. Table 1 reports their demographic and clinical Demographic and clinical characteristics and technical features of the procedure of 360 patients subjected to liver transplantation
Table 1
N (%) Demographic and clinical characteristics Age (years)a Male Hepatitis B virus Hepatitis C virus Alcoholic cirrhosis Hepatocellular carcinoma Fulminant hepatitis Primary biliary cirrhosis Primary sclerosing cholangitis Cryptogenic cirrhosis Other causes Technical features of the transplantation Split liverb Total ischemia timea Repeat liver transplantation Renal liver transplantation Hepatic artery thrombosis Pretransplant bilirubina Pretransplant MELD scorea Graft steatosis > 30%
55 272 17 104 85 101 10 7 5 10 21
(48–60) (74.7) (4.7) (28.9) (23.6) (28.1) (2.8) (1.9) (1.4) (2.8) (5.8)
37 466 31 11 9 3.1 14.5 35
(10.2) (399–560) (8.5) (3.0) (2.5) (1.3–7.2) (9.0–22.1) (9.6)
MELD, Model for End-Stage Liver Disease. a Median and interquartile range. b All patients with split liver graft received a choledochocholedochostomy.
characteristics as well as transplantation technical features. All patients received a liver from a deceased donor. A total of 142 (39%) patients developed posttransplantation biliary complications and were subsequently primarily subjected to ERCP. Table 2 reports the clinical features and procedure characteristics of 142 patients who underwent ERCP. Among patients treated because of a single posttransplantation complication, the majority developed AS (n = 79, 55.6%), followed by leaks (n = 18, 12.7%) and NAS (n = 14, 9.9%). Nineteen (13.4%) patients were treated because of strictures combined with bile duct stones, whereas 12 (8.6%) developed strictures and leaks at the same time (Table 2). The 14 patients with NAS were treated by a percutaneous intervention after undergoing ERCP for diagnostic confirmation, whereas ERCP-guided endoscopic therapy was attempted in 128 (90.1%) patients. Five (3.9%) of these patients were treated using the rendezvous technique (data not shown). The fourteen patients, who developed NAS, were excluded from analysis. In total, 490 ERCPs were performed, with a median number of three ERCPs per patient (interquartile range 2–5) (data not shown). Patients with both stones and strictures had a median number of four ERCPs per patient (interquartile range 2–5), followed by 3.5 in those with leaks and strictures (interquartile range 2–7), whereas patients with only AS or leaks had a median of three procedures per patient (interquartile range 2–5 and 2–3.75, respectively) (Table 2). The maximal average of four stents per patients was placed in patients with both stones and strictures and they were changed until a median of 24 months (Table 2). Among 128 patients who underwent ERCP-guided endoscopic therapy, definitive success was achieved in 83 (65%) (data not shown). As for the specific subgroups, the best therapeutic response was found among patients with AS (68%) and worse among those with both leaks and strictures (50%) (Table 2). We recorded four (3.1%) complications in total. Three of those were in the form of postsphincterotomy bleeding in patients treated because of AS, whereas one involved perforation in a patient with post-transplantation bile leak (Table 2). All bleeding episodes responded to endoscopic treatment (two patients with a submucosal injection of adrenaline solution and one patient with hemoclip positioning) and the perforation case responded to conservative treatment, without the need for surgical management. Predictors of outcome of endoscopic retrograde cholangiopancreatography-guided endoscopic treatment
Table 3 reports the technical features of transplantation surgery and endoscopic procedure associated significantly with the outcome of ERCP-guided endoscopic treatment (Hosmer–Lemeshow’s goodness of fit P = 0.64). Patients with a pretransplant MELD score more than 10 (OR 3.88;
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Endoscopic outcome in liver transplantation Faleschini et al. 153
Clinical features and treatment characteristics of 142 patients who underwent endoscopic retrograde cholangiopancreatography because of post-transplantation biliary complications
Table 2
Type of complications
N
Number of procedures (median)
Number of stents (median) (months)a
Stents retention time (median) (months)
Endoscopic success (%)
Complications
Anastomotic stricture Nonanastomotic stricture Leak only Leak + stricture Stones + stricture
79 14 18 12 19
3 1 3 3.5 4
3 NA 2 3 4
10.5 NA 10 11.5 24
68 NA 61 50 63
3 (bleeding) NA 1 (perforation) 0 0
a
Refers to the total number of stents inserted during the entire treatment period.
Transplantation surgery and technical features of the endoscopic procedure significantly associated with failure of endoscopic retrograde cholangiopancreatography-guided endoscopic therapy in 128 patients treated because of post-transplantation biliary complications
Table 3
OR (95% CI)a Pretransplant MELD score > 10 Stents retention time > 12 months Procedure of dilatation and stenting Stents diameter of 10 Fr
3.88 6.45 0.10 0.21
(1.16–12.95) (2.14–19.42) (0.03–0.30) (0.07–0.70)
P value 0.03 < 0.01 < 0.01 0.01
CI, confidence interval; MELD, Model for End-Stage Liver Disease; OR, odds ratio. Adjusted for recipient sex and age, indication for transplant, type of biliary complication, total ischemia time, Model for End-Stage Liver Disease. a
95% CI 1.16–12.95), as well as those who had a stent retention time of more than 12 months (OR 6.45; 95% CI 2.14–19.42), were less likely to respond favorably to endoscopic therapy. However, patients subjected to both the stenting and dilatation procedures (OR 0.10; 95% CI 0.03–0.30), as well as those treated with 10 Fr stents were more likely to achieve resolution of the complications (OR 0.21; 95% CI 0.07–0.70). On univariate analysis, hepatitis C virus infection appeared not to be associated with endoscopic outcome (P = 0.70).
Survival analyses
Of the 128 patients who were treated by ERCP in this study, none was lost to follow-up. Overall estimated 1-, 3-, and 5-year patient survival was 92, 82, and 77%, respectively (data not shown). In December 2012, 93 patients were still living, with a median follow-up from LT of 5.1 years (interquartile range 2.1–8.1 years) and 4.2 years (interquartile range 1.9–6.9 years) from first ERCP. Table 4 Factors associated significantly with survival and mortality in 128 patients treated with endoscopic retrograde cholangiopancreatography-guided endoscopic therapy because of post-transplantation biliary complications
Time to complication > 3 months Number of stent placed ≥ 5 Hepatic artery thrombosis
HR (95% CI)a
P value
0.24 (0.10–0.56) 0.31 (0.12–0.79) 13.88 (4.08–47.25)
< 0.01 0.02
