Objectives: Liver transplant represents the criterion standard therapy for end-stage liver disease. Biliary complications after liver transplant have shown an increased trend and are characterized by anastomotic and nonanastomotic stenoses.
Material and Methods: This retrospective single-center observational study included 217 patients who under-went liver transplant between January 2004 and December 2014; 18 patients had anastomotic (8.3%) and 29 (13.4%) had non-anastomotic stenoses. Patients with and without biliary stenosis were compared with regard to their preoperative, intraoperative, and postoperative parameters and donor characteristics. Patients with biliary stenosis were divided into 3 cohorts according to the type of endoscopic treatment performed (single plastic, multiple plastic, and fully covered self-ex-pandable metal stents). We compared the patients with different types of endoscopic biliary drainages for length and type of stenosis, presence of stones, time of onset and treatment, number of procedures, complications, and success rate.
Results: Preoperative Child-Pugh and Model for End-Stage Liver Disease scores, complication and re-operation rates, and donor age were significantly higher in the stenosis group. We found no statistical differences other than length of stenosis between patients with multiple stents and self-expanding metal stents.
Conclusions: Preoperative recipient conditions and postoperative morbidities may represent risk factors for development of biliary strictures. Consequently, the optimal endoscopic treatment should be tailored to the type and the onset of stenosis and the patient’s condition.
Key words : Biliary complications, Biliary stent, Endoscopic biliary drainage, End-stage liver disease
Major advances have been made in liver trans-plantation over the past decades. Novel technologies have resulted in more accurate surgical inter-ventions, enabling better outcomes associated with more efficient immunosuppressive therapy. Recently, despite overt improvements, donor-recipient pool mismatches due to organ shortages have led to the consideration of marginal grafts for transplant. As a consequence, more demanding postoperative courses and higher numbers of complications have been recorded. Biliary stenosis, leaks, stones, and Oddi dysfunction are all complications observed after liver transplant, with rates ranging from 22% to 64%. Such complications have resulted in significant morbidity, graft loss, and mortality rates,1-6 as well as economic implications.7 Such detrimental effects require prompt therapeutic decisions to reduce the risk of graft loss and poor patient survival.
In the present study, we focused on biliary stenosis posttransplant and its management. Biliary stenosis can be classified as anastomotic stenosis (AS) and nonanastomotic stenosis (NAS). The incidence of AS is between 5% and 10%, whereas NAS ranges between 5% and 25%.8,9 Anastomotic stenosis is located between the donor and recipient bile duct; it is identified by fibrotic short narrowing localized at the anastomosis site, as a consequence of fibrotic healing after ischemic damage. Such damage usually occurs within the first year posttransplant and is usually related to the surgical procedure, including technique, material, tension, and excessive use of electrocautery.10 Nonanastomotic stenosis can be single or multiple and may involve any site of the biliary tree, including extrahepatic (ie, hepatic duct confluence) and intrahepatic. Most NAS (80%) are localized around or below the bifurcation of the common bile duct, are characterized by early presentation (in the first months after transplant), and are related to hypoperfusion and ischemia of the extrahepatic tract. Late NAS more frequently involves small and peripheral bile ducts and is more often associated with immunologic factors.11 Therefore, NAS is commonly subdivided into macroangiopathic types, which occur after hepatic artery9 thrombosis, and microangiopathic types, which are usually associated with a patent hepatic artery. Micro-angiopathic types are considered “ischemic-type biliary lesions” due to many different causes related to recipient and donor characteristics, organ storage and preservation, ischemia-reperfusion injury, surgical technique of bile duct anastomosis, cytotoxic bile salt, immunologic (AB0 incompatible, rejection) factors, and infective features.12-14
Management of biliary stenosis has changed over the past years, moving from a surgical approach, through percutaneous treatment, to an endoscopic route, which is becoming the first choice of treatment. Furthermore, innovative endoscopic stents have been recently introduced, increasing the therapeutic options.
In this retrospective study, our aim was to analyze factors that can determine biliary stenosis and the efficacy of endoscopic treatment. We compared 3 different strategies: single plastic, multiple plastic, and fully covered self-expanding metal stents (SEMS).
Materials and Methods
Patients and methods
Medical records of 235 patients who underwent 242 deceased-donor liver transplant procedures between January 2004 and December 2014 at the Department of General Surgery and Liver Transplantation of the University Hospital of Verona were retrospectively reviewed. Nineteen transplant procedures were not eligible for study inclusion for the following causes: 10 patients who had hepatojejunostomy procedures performed as first choice reconstruction, 4 patients referred for fulminant hepatitis, 3 patients with postoperative deaths within 30 days, and 2 patients with postoperative bile peritonitis who underwent hepatojejunostomy. Therefore, our analyses considered 223 liver transplant procedures in 217 patients (6 patients had retransplants). All patients were older than 18 years, and all liver transplant procedures were AB0 compatible.
Organ procurement and preservation were per-formed according to standard technique using low-viscosity solution; the biliary tract was flushed with cold saline solution. The preferred implantation technique was performed with preservation of inferior vena cava with sequential revascularization (cava-porta-artery). End-to-end, duct-to-duct bile anastomosis without T tube was the favored method using a 6/0 or 7/0 polydioxanone running suture.
Patients were postoperatively provided with a double immunosuppressive strategy that included tacrolimus and prednisolone. If required, a mam-malian target of rapamycin inhibitor (everolimus) was also used in our immunosuppressive strategy. An early introduction was considered in case of renal insufficiency and hepatocellular carcinoma (HCC), especially in the event of extended HCC in explanted livers. A late indication for everolimus introduction included renal impairment, HCC recurrence, or de novo neoplasia.
Doppler ultrasonography was performed routinely on postoperative days 1, 2, and 5 and on demand to rule out vascular complications. After patients were discharged, they were followed weekly in the outpatient clinic for the first month postsurgery, monthly for 3 months, and thereafter every 4 months during the first 2 years and from then on annually. If biliary complications were suspected, magnetic resonance cholangiopancreatography was performed, and endoscopic retrograde cholangiography (ERC) with occlusive balloon was carried out when stenosis was suspected or when demonstrated with magnetic resonance cholangiopancreatography.
At the beginning of the present study period, percutaneous transhepatic cholangiography (PTC) was the first choice procedure in early AS (2-4 mo after liver transplant) because endoscopic access was considered highly risky for biliary anastomosis. Between 2004 and 2011, 15 of 24 patients underwent PTC; after 2011, PTC was performed only in 6 of 23 patients. Nevertheless, all patients subsequently underwent endoscopic management of the stenosis with placement of stents and were therefore included in the present study.
Biliary stenoses were commonly classified according to the grade of stricture. Patients were thus stratified into 3 different classes: low grade (lumen narrowing < 50%) in 15 patients (31.9%), high grade (lumen narrowing > 50%) in 15 patients (31.9%), and severe stenosis (when the stenotic tract can be passed only through hydrophilic guide wire) in the remaining 17 patients (36.2%). The choice of the endoscopic treatment had been determined according to the type of stenosis (AS or NAS), the site, and the time of onset (early or late). When treatment was not successful, biliodigestive surgery or retransplant was considered.
The study was performed in accordance with the principles of the Helsinki Declaration. All patients provided written consent to participate and to have data collected.
Descriptive statistics are presented as means and standard deviation or median with range for continuous variables and as numbers with percentages for categorical variables. Differences between groups were determined with chi-square and Kruskal-Wallis tests for discrete variables and t test and analysis of variance for continuous variables. Survival analysis was performed using the Kaplan-Meier method. Significance was defined as P < .05. Statistical analysis was performed with STATA 11.0 software (StataCorp, College Station, TX, USA).
During the study period, 223 liver transplant procedures were performed in 217 patients. Patient characteristics are listed in Table 1. Six patients underwent retransplant due to hepatic abscess (1 patient), recurrence of primary sclerosing cholangitis (1 patient), posttransplant hepatic failure (1 patient), secondary cirrhosis related to hepatitis C virus recurrence (1 patient), and chronic rejection (2 patients). The leading causes of cirrhosis were alcoholic and viral. Other than advanced history of cirrhosis and donor age, patients with biliary stenosis versus those without stenosis showed poor postoperative course with high incidence of complications (53.1% vs 30.6%; P = .004), increased grade III and IV disease according to Dindo-Clavien classification (P = .01),15 and a higher reoperation rate (34% vs 13.6%; P = .001). Preoperative clinical history showed that patients with biliary stenosis had significant higher Child-Pugh class of cirrhosis and increased Model for End-Stage Liver Disease (MELD) scores (17 vs 20.1; P = .02).
Patients without biliary stenosis (171 patients/total number of 176 transplant procedures) were compared with patients with biliary stenosis (46 patients/47 transplant procedures). Median follow-up was 55 months (range, 17-135 mo). Overall survival in patients without versus patients with biliary stenosis at 3, 5, and 10 years was 90%, 86%, and 71% versus 81%, 71%, and 60% (not significant). The 47 patients with biliary stenoses (21% of total transplants) were divided in 18 with AS (8%) and 29 with NAS (13%). Nonanastomotic stenosis was extrahepatic in 12 cases, at the biliary confluence in 7 cases, and intrahepatic in 10 cases. Survival rates at 3 and 5 years in AS versus NAS were 81% and 70% versus 83% and 60% (not significant), respectively (Figure 1). A significantly shorter stenosis was measured in the AS group (7.3 vs 21.5 mm; P < .001). No statistical differences were observed for early postoperative parameters (such as acute arterial thrombosis, acute rejection, and cytomegalovirus infection) or T-tube insertion (Table 2).
Median length of stenosis was 14.5 mm (range, 4-40 mm). Patients with biliary stenosis were divided into 3 groups according to the endoscopic treatment performed: 22 patients underwent single plastic stent positioning (7 AS and 15 NAS) (Figure 2), 11 patients underwent multiple plastic stents (2 AS and 9 NAS) (Figure 3), and 14 patients underwent SEMS (9 AS and 5 NAS) (Figure 4).
When we analyzed relationships among these 3 groups of patients, a significant difference (P = .043) was shown regarding the number of patients with AS versus NAS. When we considered grade of stenosis in each group, no differences were noted (Table 3). Survival rates with low-, high-, and severe-grade stenosis in the single stent group at 3 years were 71%, 85%, and 85%; at 5 years, rates were 57%, 22%, and 83%, respectively. Survival rates in the multiple stent group were 75%, 100%, and 100% at 3 years and 75%, 100%, and 100% at 5 years, respectively. Respective survival rates in the SEMS group were 100%, 100%, and 66% at 3 years and 50%, 100%, and 66% at 5 years (not significant).
Clinical success was considered as complete resolution of symptoms in the
absence of documented stenosis on imaging studies. In our study, we found a
lower success rate compared with rates shown in the literature.3,16-18 A
possible explanation could
be correlated with patients currently undergoing endoscopic procedures with indwelling stent. Survival (Kaplan-Meier) rates for single stent, multiple stent, and SEMS groups at 3 years were 78%, 85%, and 85%. At 5 years, rates were 52%, 85%, and 72%, respectively (Figure 5). Overall survival rates in the low-, high-, and severe-grade stenosis patients at 3 and 5 years were 76%, 93%, and 80% and 60%, 41%, and 80%, respectively (not significant).
When we overlooked the type of stenosis, we found differences between groups of treatment (Table 3). Different procedure outcomes and complications were strictly related to the duration and/or number of surgical procedures. The lapse of time between diagnosis and liver transplant in the single stent group was shorter (median of 5.5 months) than the other groups (median of 8.9 months), as, with early stenosis, it is preferred to avoid dilatation and to place a single stent. Mean length of stenosis showed a significant difference between multiple stent (18 mm) and SEMS (8.5 mm) (P = .02) because the SEMS group had more patients with AS, which has been demonstrated to be shorter than NAS.
Complications included bleeding, pancreatitis, and cholangitis, which were similar within the 3 groups of patients. Notably, sepsis events within the single stent group resulted in death in 2 of 22 patients (9%).
In the SEMS group, displacement of the stent, originally placed transpapillary with the tip in the duodenum upward in the distal common bile duct, was observed. Nevertheless, the stent could be managed and removed endoscopically without complications. In addition, the SEMS procedure resulted in a higher success rate (57%) than use of single (41%) or multiple (27%) stents, and no patients were referred to surgery due to treatment failure. However, 3 patients died due to graft loss secondary to biliary cirrhosis. In the multiple stent group, treatment was successful and patients had a reduced number of procedures and shorter procedure times.
Preoperative recipient conditions may contribute to biliary stenosis, as highlighted by preoperative bilirubin levels that correlate with posttransplant outcomes, graft survival, and arterial thrombosis risk.1 It has been almost 20 years since the first report that associated poor Child-Pugh status with increased incidence of NAS.19,20 Our study also supported this scenario in which Child-Pugh and MELD scores efficiently predicted biliary stenosis. However, it cannot be ruled out that such outcomes may be also the consequence of an increased use of organs from extended-criteria donors from pre- to post-MELD era.21 Transarterial chemoembolization is another preoperative risk factor in the development of biliary stenosis in patients with hepatocellular carcinoma.16 The mechanism of damage is determined by vasculitis due to direct trauma to vessels, local toxicity from therapeutic agents, and systemic immune response.22
Mammalian target of rapamycin inhibitors may have an active role in terms of fibrosis, which may be associated with early recurrent biliary obstruction.23,24 Although an association between mammalian target of rapamycin inhibitors and biliary stenosis is a possibility, we cannot confirm this due to our heterogeneous study population. In fact, our study comprises a long period of time with different immunosuppressive strategies, which may have varied over time after transplant. In addition, everolimus monotherapy represents, in our popu-lation, an infrequent treatment strategy (about 1% of the total population), as most patients received a combined therapy of calcineurin inhibitors and everolimus. This variability introduced a con-founding bias, limiting us from reaching any definitive conclusions.
Recent evidence has encouraged the use of choledochocholedochostomy, a technique that allows an endoscopic approach, preserves the Oddi sphincter as a natural barrier to bacterial reflux,2 and has a lower incidence of complications than hepaticojejunostomy.25 Nevertheless, splinting a biliary anastomosis with a T tube is still matter of discussion, as complications related to T-tube removal can lead to further morbidity.2
Liver transplant is a complex surgical procedure often performed in patients with poor health conditions, which is characterized by a high risk of postoperative complications. In our study, patients with biliary stenosis versus those without had significantly higher rates of overall complications (53.1% vs 30.6%; P = .004), but no differences were noted regarding cold or warm ischemia time and type of biliary reconstruction. Complications, classified according Dindo-Clavien score,15 showed significantly elevated life-threatening complications in the biliary stenosis group (P = .01). Several complications, including bleeding, hematoma, hepatic artery thrombosis, biloma, and biliary leaks require surgical procedures as highlighted by significantly higher reoperation rates in the biliary stenosis group (34% vs 13.6%; P = .001). A possible correlation between postoperative complications that were surgically treated and biliary stenosis might be the need to mobilize the graft during reoperation in an effort to drain, control bleeding, or redo arterial anastomosis, which can lead to minor unrecognized disruption of the choledochocholedochostomy.
Over the previous decades, the endoscopic approach has become the preferred procedure in patients with duct-to-duct anastomosis.26 Its indication largely depends on several factors, including whether stenosis onset is early or late posttransplant, the type (AS, NAS) and site of stenosis (extrahepatic, confluence, intrahepatic), and patient condition. The optimal treatment should be tailored to the individual patient, including the possibility to perform balloon dilatation and the ability to place single stents, multiple stents, or SEMS. Although balloon dilatation is commonly avoided in the early AS to prevent the risk of disrupting the anastomosis,9 it could be used in short stenosis; in such cases, balloon dilatation alone has a success rate of around 41%, although recent data have estimated a high success rate (74%) when associated with stent insertion.27
The duration of single stents is poor; nevertheless, it is indicated in small-caliber ducts (extrahepatic or intrahepatic) that do not allow for a second stent and in early strictures without balloon dilatation.17 It has also been used as a bridge to more definitive therapy, especially in septic patients who need immediate drainage without prolonged procedures.28 The poor survival results shown in our single stent group at 5 years (52%) could be related to a larger number of patients with NAS (15 patients) with multiple and intrahepatic stenoses, which is a critical condition for patients.
The criterion standard for extrahepatic biliary stenosis after liver transplant is the use of multiple stents; use of multiple stents has shown good duration and good long-term results. The rationale of multiple stent placement is to maintain the maximal luminal size achieved during balloon dilatation, promoting remodulation of bile ducts over the stents and preventing duct narrowing while stents are still in place.18,29 Furthermore, multiple stents may reduce complications related to stent occlusion, such as jaundice and cholangitis, by adding biliary drainage through inter-stent spaces. The main limitation of this method is the need for multiple ERC to replace the stents and to add more when possible. Its success rate is between 92% and 100%.30 However, in our study, the success rate of multiple stents was actually low due to the presence of preexistent stents in many of the patients. Although patients with multiple stents had good overall conditions, they cannot be considered cured since most had NAS, which has a lower success rate with multiple stents than AS.8,31
The possibility to achieve a larger size dilatation with an opportunity for removal is the appealing advantage of partial and fully covered SEMS.32-35 Since its introduction in benign biliary stenosis in 2008, SEMS has gained use, as it is safe, represents an option in refractory strictures, and can be left in place for longer periods of time. Its major limitations are upward migration and inability to remove the stent due to embedding.36 However, newly designed stents with anchoring flaps or flared ends and drug-eluting stents have been introduced in clinical practice, with promising results in terms of lower migration rate and prolonging stent patency.37 Compared with multiple stents, SEMS has a similar stenosis resolution rate (90%).30,32,38-40
Our survival analyses showed no significant differences between patients with AS versus NAS, among the 3 different endoscopic treatment groups (single stent, multiple stent, SEMS), and the grade of stenosis. However, conclusions are limited due to small sample sizes.
The therapeutic endoscopic choices can be summarized as follow. For early AS (within the first 2-4 mo after liver transplant), sphincterotomy, no dilatation, and 10F single stents are preferred. For late AS, before the introduction of SEMS, sequential multiple stents should be used, starting first with a single stent and then ERC every 3 months, with insertion of 1 more parallel stent each time (up to 4 or 5 stents) without removal of the previous stent if still patent. The time limit for treatment decisions is determined by recurrence of symptoms.
After the introduction of SEMS, sphincterotomy, dilatation, and 10-mm SEMS should be used. For patients with NAS, extrahepatic stenosis is treated as late AS. Hilar stenoses are treated with multiple stents in the effort to drain every hepatic segment. Intrahepatic stenosis of secondary order ducts is managed with dilatation and single stents due to the small size of the duct.
Our study confirmed that poor preoperative patient status (Child-Pugh and MELD scores) and donor age, which can affect graft function, may influence the risk of developing biliary stenosis. Although we recognize the possible bias regarding ongoing treatment of some patients during the study period and the small number of patients, we observed no differences regarding intraoperative findings. However, postoperative complications (bleeding, hematoma, biloma, biliary leak, and hepatic artery thrombosis requiring reoperation) were higher in patients with biliary stenosis, underlining the possibility that reoperation itself may be a risk factor in determining biliary complications.
Therapeutic options for biliary stenosis after liver transplant now include endoscopic access, thanks to the development of new accessories and endoscopic armamentarium. Depending on the experience of the endoscopist, most duct-to-duct biliary stenoses can be treated endoscopically either with single stents, multiple stents, or SEMS, which is dictated per the site of the stenosis and the patient’s condition. Although we recognize the possible bias regarding ongoing treatment of some patients, we found that no method was superior to the other. Thus, the best procedure should be tailored to the patient, with short intrahepatic stenosis better treated with a single stent and SEMS more for an anastomotic stenosis. Long extrahepatic or hilar strictures are better treated with multiple stents. A multidisciplinary evaluation is mandatory to determine the best treatment for the patient’s biliary stenosis.
Although care must include improving the patient’s condition before liver transplant, it is important to develop better tools to assess graft quality, including biomarkers to detect cellular injury,41 in the era of organ shortages where machine perfusion for optimization of grafts from donors after circulatory death has been acquiring more weight.
DOI : 10.6002/ect.2018.0167
From the 1Department of General Surgery and Liver Transplantation, the
2Department of General Surgery, Surgical Endoscopy, and the 3Department of
General Surgery, General and Upper GI Surgery, University Hospital of Verona,
Verona, Italy; the 4Department of Pathology, University of Verona, Verona,
Italy; the 5Department of Laboratory Medicine, Division of Pathology, Karolinska
University Hospital, Stockholm, Sweden; and the 6Chief Medical Office,
University Hospital of Verona, Verona, Italy
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Luca Bortolasi, Department of General Surgery and Liver Transplantation, AOUI Verona, p.le Stefani 1, 37126 Verona, Italy
Phone: +39 0458121949
Figure 1. Kaplan-Meier Survival in Anastomotic Stenosis Group Versus Nonanastomotic Stenosis Group
Figure 2. Single Plastic Stent in Short Early Anastomotic Stenosis
Figure 3. Multiple Stents in Hilar Nonanastomotic Stenosis
Figure 4. Fully Covered Self-Expandable Metal Stent (SEMS) in Late Anastomotic Stenosis
Figure 5. Kaplan-Meier Survival in Single Stent Versus Multiple Stent Versus Fully Covered Self-Expandable Metal Stent Groups
Table 1. Patient Characteristic in Groups With and Without Biliary Stenosis
Table 2. Postoperative Complications in Patients With Anastomotic Stenosis and With Nonanastomotic Stenosis
Table 3. Comparisons of Different Endoscopic Approaches