Percutaneous Transhepatic Biliary Drainage for Biliary Stricture After Endotherapy Failure in Living Donor Liver Transplantation: A Single-Centre Experience from India

Abstract

Background

In the living donor liver transplant (LDLT) population, postoperative biliary stricture is a common problem. Endoscopic retrograde cholangiography (ERCP) and stenting failures usually occur because of odd angulation of the ducts, particularly in right lobe grafts. Percutaneous transhepatic biliary drainage (PTBD) is helpful in patients where endoscopic intervention is unsuccessful. We report our experience in this particular group of patients from India.

Methods

We analysed retrospective data of 932 patients who underwent LDLT at our centre. Data collected include basic demographics, technical and clinical success rate of PTBD, postprocedure complications and overall outcome.

Results

Of the 932 patients, 81 (8.7%) developed biliary stricture and required endoscopic or percutaneous interventions; out of whom, 75 patients underwent ERCP, and 6 patients had hepaticojejunostomy. Right posterior sector graft recipients had the highest incidence of biliary stricture (29.16%), followed by the right lobe grafts (8.57%), whereas stricture rate was less among left lobe and left lateral lobe graft recipients (3.77% and 3.7%, respectively). Failure of ERCP was common among the ductoplasty group (80%). The failure rate of ERCP did not differ significantly among patients who had single-duct, two-duct or three-duct anastomoses (44%, 36% and 50%, respectively, p > 0.05). PTBD was technically successful in 34 patients (87%) The mortality rate following PTBD was 2.5%, and morbidity was 15%. For the 5 patients in whom PTBD could not be performed, open hepaticojejunostomy was performed successfully.

Conclusions

Right lobe liver grafts have high incidence of biliary stricture compared with left lobe grafts. The failure of ERCP was not dependent on the number of biliary anastomoses; however, the ductoplasty group had a higher failure rate. PTBD is technically successful in a majority of patients albeit with a morbidity of 15%. Open hepaticojejunostomy is a good therapeutic option following PTBD failure.

Biliary complications including stricture and leak are an important cause of morbidity and mortality in living donor liver transplantation (LDLT) recipients.¹ Biliary morbidity constitutes the second most common cause of graft dysfunction following rejection.² The incidence of biliary complications is higher in LDLT recipients (14–32%), especially under the form of biliary leakage from the resection margin of the graft.³ A recent systematic review revealed an overall incidence of 13% and 8% for stricture and leak, respectively.⁴ In contrast to deceased donor liver transplantation, there are multiple factors in the setting of partial liver transplantation which increase the risk of biliary complications. The smaller size of the ducts, number of ducts which is often more than one and right versus left lobe grafts with inherent anatomical differences are unique factors affecting biliary outcomes to an LDLT setting.⁵ These complications generally require therapy by endoscopic percutaneous radiologic procedures or less commonly require surgical interventions.⁶ In this present study, we present data of utilisation of each modality with particular reference to percutaneous biliary interventions from our experience.

Materials and methods

After obtaining the institutional review board and ethics committee approval, we have collected the retrospective data of all patients who have undergone LDLT at our centre between September 2012 and October 2015. The 1-year mortality rate during this time period was 9.35%. The remaining patients were on regular follow-up according to the unit protocol. The last follow-up was in November 2015. Biliary complications occurring as a result of hepatic artery thrombosis, nonanastomotic biliary stricture, patients who developed bile leaks in the early postoperative period and did not require any long-term management and deceased donor liver transplantation (DDLT) were excluded from the analysis. The remaining patients with biliary complications who had either bile leak or biliary stricture and required intervention were included in the study. Basics demographics, clinical data, technical and clinical success rate of percutaneous transhepatic biliary drainage (PTBD), postprocedure complications and short-term outcome details were collected from the medical records.

The diagnosis of biliary stricture was made based on classic clinical symptoms such as jaundice and itching in association with elevated bilirubin and alkaline phosphatase/gamma-glutamyl transferase levels. Some patients presented with the features of cholangitis including fever, raised leucocyte counts and transaminase levels. All patients suspected to have biliary stricture were routinely evaluated with magnetic resonance cholangiography (MRC) scan. By delineating the number, site and extent of strictures, an MRC scan provides a roadmap for further therapeutic interventions particularly in situations where there are multiple biliary anastomoses. On magnetic resonance imaging (MRI), biliary strictures appear as a tight, smoothly marginated focal zone of decreased signal intensity in the line of bile ducts with upstream biliary dilatation.³ However, MRC is not sensitive or specific for diagnosis of biliary stricture in the transplanted liver. We have encountered situations where a patient presents with characteristic clinical and biochemical features with a normal appearing duct on MRI. This can be due to reduced bile duct compliance secondary to graft-related factors.⁷

An informed written consent was obtained from the patient or legal guardian, before withholding immunosuppression on the day of percutaneous transhepatic procedure. At our center, percutaneous transhepatic procedure was preferred as first-line therapy in patients with bilioenteric anastomoses. Endoscopic intervention was the procedure of choice for strictures arising from a duct-to-duct reconstruction (Figure 1). In cases of endoscopic procedure failure, patients were subjected to percutaneous procedure. Endoscopic failure was defined as the inability to pass a guidewire through a tight common duct stricture or failure to cannulate one or both ductal systems (anterior and posterior ducts in a right lobe graft). For strictures arising from hepaticojejunostomy, patients were subjected directly to percutaneous interventions as endoscopic access is difficult because of altered anatomy.

Figure 1.

Schematic representation of our treatment protocol for biliary stricture. ERCP, endoscopic retrograde cholangiography; PTBD, percutaneous transhepatic biliary drainage.

Percutaneous Transhepatic Biliary Drainage

All PTBD procedures were performed by the senior interventional radiologist under general anaesthesia. Ultrasound-guided, peripheral or central intrahepatic bile duct was punctured with an 18-gauze needle in dilated biliary systems. In undilated biliary ducts, which are often the case in an LDLT setting, a 22-gauge Chiba needle (Cook Medical Inc, Bloomington, USA). A 0.018-inch angled hydrophilic guidewire (Terumo, Tokyo, Japan) was passed, and the Chiba needle was exchanged for a 5-Fr coaxial dilator and a 6-Fr to 7-Fr sheath. A Kumpe catheter (Cook Medical Inc, Bloomington, USA) or a 5-Fr cobra catheter (Cook Medical Inc, Bloomington, USA) was used to traverse the stricture. Once the stricture was traversed, it was dilated with 6–8-mm balloons (Cronus II; Medispes, Zug, Switzerland or Synergy; Boston Scientific, Galway, Ireland). Two to 3 dilatations were performed with each session lasting 1–3 min until the balloon waist is eliminated. An internal-external drainage catheter (Cook Medical Inc, Bloomington, USA) was then placed across the stricture with the distal tip reaching the duodenum. The catheter was left for drainage in a closed bag system to observe for transient hemobilia. Following the percutaneous procedure, 24–48 h later, the patient was subjected to for endoscopy and subsequent internalisation.⁸ Patients who had frank cholangitis and those who could not undergo internalisation were left with percutaneous catheters. These external catheters were capped for internal drainage at discharge.

Follow-up

Complete liver function test (LFT) was performed 3 days after discharge from the hospital. Subsequent LFTs were carried out at monthly intervals. Biliary stent exchange was performed at 3-month intervals or earlier if the patient manifested signs of stent blockage and cholangitis. Ursodeoxycholic acid was added to the medication list including immunosuppressive agents as per our unit protocol. For patients who underwent PTBD, ‘technical success’ was defined as successful placement of the prosthesis across or above all biliary strictures. ‘Clinical success’ was defined as subsequent normalisation of LFTs and improvement of symptoms. For patients with long-term indwelling biliary drainage catheters, we routinely changed catheters every 8 weeks to prevent catheter clogging. Biliary drainage catheters were finally removed when follow-up laboratory data showed stable serum bilirubin and liver enzymes and subsequent cholangiography revealed fluent passage of contrast media without bile leaks and with residual stenosis less than 50% of the normal extrahepatic bile duct. If there was evidence of restenosis, however, the drainage catheter was repositioned across the stricture for an additional 3 or more months following repeat balloon dilation. Following drainage catheter removal, patient's symptoms and biochemical data were routinely assessed every 6–12 weeks at the liver transplant outpatient clinic.

Statistical Analysis

The data were analysed using SPSS version 20 (IBM Corp. Released 2011, Armonk, NY) and Microsoft Excel. Standard statistical methods for assessment on proportions, percentages and measures of central tendencies (mean, standard deviation, median and range) were used. Continuous variables were expressed as mean or median with an interquartile range; categorical variables were expressed as frequency and percentage.

Results

A total of 932 patients' data were included in the final analysis. Among 932 patients, 828 (89%) received right lobe liver grafts, 24 (2.5%) received right posterior sector grafts, 53 (5.7%) received whole–left lobe grafts and 27 (2.9%) received left lateral segment grafts (Table 1). Eighty-one patients developed biliary stricture resulting in an overall incidence of 8.7%. Right posterior sector graft recipients had the highest incidence of biliary stricture (29.2%), followed by the right lobe grafts (8.7%), whereas stricture rate was less among left lobe and left lateral lobe graft recipients (3.8% and 3.7%, respectively, Table 2).

Table 1.

Graft Types and Stricture Rates.

Type of graft	No. of patients	Biliary stricture (n)	Biliary stricture rate (%)
Right lobe	828	71	8.57
Left lobe	53	2	3.77
Left lateral segment	27	1	3.70
Right posterior sector	24	7	29.16

Table 2.

Graft Types and Intervention.

Type of graft	ERCP	PTBD
Right lobe	67	36
Left lobe	1	1
Left lateral segment	0	1
Right posterior sector	7	1
Total	75	39

Failed ERCP and hepaticojejunostomy patients proceeded to PTBD.

ERCP, endoscopic retrograde cholangiography; PTBD, percutaneous transhepatic biliary drainage.

Of the 81 patients who developed biliary stricture, 75 patients underwent endoscopic retrograde cholangiography (ERCP), and 6 patients had hepaticojejunostomy proceeded with PTBD (Figure 2). Of the 75 patients who underwent one or more ERCP procedures, 33 patients (44%) subsequently underwent PTBD. Biliary anastomosis details in each group are presented in Table 3. Failure of ERCP was due to the inability to pass the guidewire in one of the two (anterior/posterior) ductal systems in 22 cases and tight impassable common duct stricture in 9 cases, and in 2 patients, ERCP was abandoned because of bleeding. Failure of ERCP was common among the ductoplasty group (80%, Table 4). The failure rate of ERCP did not differ significantly among patients who had single-duct, two-duct or three-duct anastomoses (44%, 36% and 50%, respectively, Table 4).

Figure 2.

Clinical data of 81 patients. ERCP, endoscopic retrograde cholangiography; PTBD, percutaneous transhepatic biliary drainage.; HJ, Hepaticojejunostomy.

Table 3.

Biliary Anastomosis Type and Intervention.

Biliary anastomosis type	ERCP (n = 75)	PTBD (n = 39)
Single duct	43	19
Ductoplasty	5	4
2 ducts	25	9
3 ducts	2	1
Hepaticojejunostomy	0	6

ERCP, endoscopic retrograde cholangiography; PTBD, percutaneous transhepatic biliary drainage.

Table 4.

Biliary Anastomosis Type and ERCP Failure.

Biliary anastomosis type	ERCP failures (%)
Single duct	19 (44%)
Ductoplasty	4 (80%)
2 ducts	9 (36%)
3 ducts	1 (50%)

Overall failure rate is 44% with no statistically significant difference in ERCP failure among these groups, p = 0.34.

ERCP, endoscopic retrograde cholangiography.

PTBD was technically successful in 87% of patients (34 of 39). Balloon dilatation and subsequent internalisation were performed in 29 patients, whereas the other 5 patients had only external drainage (Figure 3). The mean bilirubin value was 7.28 ± 6.2 (range, 1.70–27.0) mg/dL and 3.86 ± 3.5 (range, 0.80–13.0) mg/dL, before and after PTBD, respectively (p < 0.05). Six (15%) patients developed complications: cholangitis (n = 2), gastrointestinal bleed (n = 2) necessitating angioembolization in one patient, hypotension (n = 1) and thoracic complication necessitating thoracotomy (n = 1). There was one death reported following PTBD because of cholangitis, sepsis and subsequent GI bleed. Open hepaticojejunostomy was performed successfully for 5 patients for whom PTBD was not possible.

Figure 3.

Clinical data of the PTBD cohort. PTBD, percutaneous transhepatic biliary drainage.

Discussion

Biliary complications constitute the most important cause of morbidity for a LDLT recipient and are regarded the ‘Achilles heel’ of LDLT.⁹ The overall incidence of biliary complications in LDLT recipients varies from 5% to 40% and is definitely higher than the DDLT recipient population.¹⁰ Most biliary strictures are diagnosed and managed by nonoperative means including endoscopic and percutaneous procedures. ERCP is regarded as the first choice for management of biliary strictures, failing which patients are subjected to percutaneous procedures.¹¹ It has been shown that LDLT recipients with biliary stricture have less favourable response to endotherapy, thereby requiring percutaneous interventions more often than the nontransplant population.¹² Several studies have retrospectively analysed the utility and outcome of therapeutic interventions in the management of biliary procedures.^{13, 14} This study presents our experience in managing patients with biliary stricture following LDLT with particular reference to patients undergoing PTBD. To the best of our knowledge, our cohort data of 39 patients who underwent PTBD for post-transplant biliary stricture are the first largest single-institution experience to be reported.

Multiple factors contribute to biliary leaks and strictures in an LDLT setting including the operating centre volume/expertise, aberrant biliary anatomy resulting in multiple small bile ducts, technical factors causing ductal ischaemia during hilar dissection^{15, 16, 17} and the number and type of anastomoses.¹⁸ In our study, we found that it was widely believed that duct-to-duct reconstruction in left lobe LDLT is prone to complications as the arterial supply to the left hepatic duct comes from the right hepatic artery and hence was generally contraindicated.^{19, 20} However, Soejima et al²¹ established the feasibility and safety of duct-to-duct reconstruction in left lobe LDLT in their series. In our series, the rate of biliary stricture is actually higher in right lobe grafts than in left lobe grafts. This is possibly due to the precise surgical techniques in hilar dissection during donor left hepatectomy.⁵ In our series, we observed a higher incidence of stricture (29%) in patients who received right posterior sector liver grafts. As the dissection happens high up in the hilum during the harvest of posterior sector grafts, the chance of biliary ischaemia is higher resulting in higher incidence of strictures. However, LDLT with posterior sector grafts is performed only in exceptional situations, and this higher incidence of biliary stricture should be taken into consideration while deciding on posterior sector grafts.

In our institution, the success rate of ERCP for biliary stricture is 56% as 44% patients had to subsequently undergo percutaneous transhepatic cholangiography (PTC). This is in accordance with a recent review which reported that the success rate of ERCP in managing strictures in LDLT recipients ranges between 37% and 71%.²² In general, the higher the number of biliary anastomoses, the higher the rate of failure of ERCP. Surprisingly, our analysis found that the ability to successfully cannulate all the sectoral ducts was not dependent on the number of biliary anastomoses. This can probably be explained by our technique of hilar dissection during recipient hepatectomy. We divided the bile duct high up at the level of the hilar plate and meticulously preserved the blood supply until that point. This ensured multiple ductal openings in the recipient side as well; thereby, we could anastomose sectoral ducts to separate openings in the recipient side. This later helped the endoscopist to pass the guidewire across the stricture even if there were multiple biliary anastomoses. Another interesting finding in our study is the higher incidence of ERCP failures in the ductoplasty group (80%). In the technique of ductoplasty, two adjacent ducts are sutured together to make it a single spectacle-shaped opening, and then, duct-to-duct anastomosis is performed. When stricture develops in ductoplasty anastomosis, the lumen is both narrowed and angulated accounting for the difficulty in passing the guidewire and the higher incidence of ERCP failures. As ductoplasty itself had higher incidence of biliary strictures, we have abandoned using this technique in our unit for biliary reconstruction.

The technical success rate of PTC in our institution is 87%. Two large studies reported a success rate of 85% and 89%.^{23, 24} Following the successful percutaneous catheterisation, our patients routinely underwent balloon dilatation and stent placement by endoscopy (rendezvous procedure). This avoided a significant morbidity of having an external catheter. Hence, our patient satisfaction was quite high despite having undergone two different procedures.

Percutaneous transhepatic biliary intervention is a major procedure and is associated with morbidity and mortality. In our series, the morbidity rate was 15%, and the mortality rate was 2.5%. Finally, there are patients who will have strictures not amenable to percutaneous treatment. In those patients, open surgical approach to access the biliary stricture and construction of Roux-en-Y hepaticojejunostomy was uniformly successful with low morbidity.²⁵

In conclusion, biliary stricture is common among LDLT recipients. Highest incidence of biliary stricture was seen in right posterior sector graft recipients (29%). ERCP is the initial procedure of choice and has a high failure rate because of technical reasons. The failure of ERCP was not dependent on the number of biliary anastomoses; however, the ductoplasty group had a higher failure rate. PTBD was technically successful in most patients albeit with a morbidity of 15%. Open hepaticojejunostomy was a good therapeutic option following PTBD failure. Although biliary complications are common among LDLT recipients, it can be successfully managed with endoscopic and percutaneous interventions.

Limitations of the study

It is a retrospective study carried out by analysis of data from hospital database. As our unit treats patients from multiple countries in the Middle East, Far East, Central Asia and Southeast Asian region, some patients might undergo endoscopic procedures in their home country itself and may not travel frequently for follow-up. Because the patients were from distant places, only short-term outcomes could be analysed in this study. Despite these limitations, we were constantly communicating with them by telephonic and electronic means to ensure adequate care.

Ethical statement

This study was approved by an independent institutional review board and ethics committee.

Conflicts of interest

The authors have none to declare.