Biliary complications after liver transplantation (LT) have always been one of the most common problems after surgery, including biliary stricture.1, 2 Anastomotic biliary strictures are thought to result from technical surgical problems or local ischemia; the incidence has been reported to be <10% for deceased donor liver transplantation (DDLT), and to be higher, up to 30%, in living donor liver transplantation.2–7 Although anastomotic strictures are frequent, they can be treated by nonsurgical procedures.8–10
Intrahepatic biliary strictures (IHBS) are included in nonanastomotic strictures; they result primarily from hepatic artery thrombosis.11–13 IHBSs, without hepatic artery occlusion, have been reported to be rare.14 Identified causes include chronic ductopenic rejection, ABO incompatibility, ischemia/reperfusion injury, or recurrence of primary disease such as primary sclerosing cholangitis or autoimmune hepatitis.15–20 Although IHBSs are less common than anastomotic strictures, they are clinically important because of their poor response to treatment. They require more frequent intervention over a longer period of follow-up compared with anastomotic strictures, and they frequently result in failed interventions that lead to retransplantation or death.1, 14, 16, 17, 21
Moreover, in non-heat-beating donor (NHBD) LT, the prolonged warm ischemic time may predispose patients to an increased incidence of ischemic type IHBSs.22 Therefore, it may be predicted that in the future, when marginal donors such as NHBDs will be used more widely because of the serious donor organ shortage, more patients will experience post-LT IHBS. Therefore, the clinical course and management strategy for IHBSs is important; however, they have not been well studied to date. The purpose of the present study was to develop a classification of post-LT IHBS and to evaluate patient prognosis to establish an appropriate management strategy on the basis of the clinical features of individual patients.
We retrospectively analyzed the medical data of 172 patients. These patients underwent DDLT in China between December 2001 and June 2005 and were followed at Seoul National University Hospital in Korea. Data from patients with early postoperative mortality in China were excluded.
Because perioperative patient data were not available, we could not determine the details of individual patient warm or cold ischemic time, type of perfusion solution, or the surgical technique used in the LT procedure. From a previous Chinese report, we assumed that most of the deceased donor organs were retrieved from uncontrolled NHBDs.23
Patients received triple-drug immunosuppression that consisted of a calcineurin inhibitor (tacrolimus or cyclosporine), mycophenolate mofetil, and prednisolone. Prednisolone was tapered gradually and discontinued at 6 months after LT. Mycophenolate mofetil was discontinued in patients who had severe gastrointestinal symptoms or neutropenia.
Most patients had a T tube placed in the common bile duct. T tube cholangiography was routinely performed right after the first visit to our center. We repeatedly evaluated patients with cholangiography to examine changes in the biliary tree if the initial cholangiography was abnormal or if biliary symptoms developed. Otherwise, cholangiography was carried out 3–6 months after LT, just before removal of the T tube. In patients without a T tube, we did not routinely evaluate the biliary tree. Either endoscopic cholangiography or magnetic resonance cholangiography was used in cases with a suspected biliary complication. The presence of an IHBS was defined when intrahepatic stricture was identified by cholangiography, regardless of the presence of biliary symptoms.
A total of 55 patients (32.0%) developed biliary complications. Among these patients, IHBS was the most frequent biliary complication (n = 44, 80.0%), followed by anastomotic stricture, major leakage, and choledocholithiasis (Table 1). Computed tomography with 3-dimensional reconstruction of the hepatic artery was performed in all patients with IHBS; no patient was found to have hepatic artery insufficiency. Thirty-nine patients (88.6%) were men and 5 (11.4%) were women. Their mean ± SEM age at the time of LT was 51.6 ± 1.2 years. The median follow-up duration after LT was 18.4 months, with a range of 3.9–43.5 months. Forty-two patients (95.5%) underwent LT for hepatitis B–related liver disease, 1 for hepatitis C–related liver disease, and 1 for alcoholic cirrhosis. The biliary reconstruction was duct-to-duct anastomosis in most patients except for one case of choledochojejunostomy. A T tube was not placed in 4 patients.
Table 1. Type of Biliary Complications (n = 55)
Intrahepatic biliary stricture
Anastomotic stricture only
Major leakage only
Interventional therapy was attempted only in patients with symptoms such as cholangitis, jaundice, or liver function abnormality. First, the T-tube tract was used for percutaneous drainage and interventions. Endoscopic drainage and dilatation were attempted next if there was no T tube or if the intervention failed through its tract. In cases of poor response to minimally invasive methods, percutaneous transhepatic drainage and dilatation were attempted. Some patients required multiple transhepatic approaches. Retransplantation was recommended in patients who had intractable IHBS that caused persistent, life-threatening cholangitis.
Classification of IHBS
According to the cholangiographic appearance of the biliary tree, we classified patients with IHBS into 4 groups: unilateral focal (UF), confluence (CO), bilateral multifocal (BM), and diffuse necrosis (DN). We defined the UF type as cases with a stricture only in the segmental branch of the unilateral hemiliver (Fig. 1); the CO type were cases with several strictures at the confluence level (Fig. 2), and the BM type were cases with multiple strictures bilaterally (Fig. 3). Cases with diffuse obliteration of peripheral ducts or destruction of the central architectural integrity, over a long segment, were included in the DN type (Fig. 4). Evaluation of biliary symptoms, necessity of and responsiveness to interventions, and prognosis were investigated for each IHBS type.
A total of 8 patients (18.2%) were included in the UF type. Their follow-up was a median of 21.3 months, with a range of 12.7–43.5 months. All patients in this group had nearly normal hepatic function and developed no biliary symptoms during the follow-up period. No patient required biliary intervention.
Ten patients (22.7%) were included in the CO type. The median (range) follow-up was 18.4 (14.0–26.2) months. Two patients (20.0%) developed intrahepatic duct stones and underwent endoscopic stone removal. Three patients (30.0%) required several interventions for biliary strictures. The interventions included endoscopic or percutaneous dilatation and drainage and temporary stenting. However, the other 5 patients (50.0%) in this group did not require any intervention. All of the patients who required several interventions had successful outcomes. Symptoms were relieved, and hepatic function was improved after the interventions. All percutaneous catheters were removed by 1 year after LT.
Twenty-one patients (47.7%) had multiple strictures throughout the entire liver and were included in the BM type. They were followed up during a median (range) of 20.8 (4.9–33.1) months. Two patients died of early cancer recurrence at 4.9 and 6.8 months after LT. Three patients (14.3%) required no interventional therapy during the follow-up period despite intermittent biliary symptoms and slightly abnormal hepatic function. Six patients (28.6%) underwent several interventional dilatations for multiple stricture sites, and the procedure resulted in a good prognosis in all patients. Their symptoms were relieved, and percutaneous catheters were removed after several interventions. However, 7 patients (33.3%) experienced recurrent cholangitis and jaundice without symptomatic improvement for >1 year after transplantation, despite repeated interventions. One patient (4.8%) died of biliary sepsis, and 2 (9.5%) underwent retransplantation for graft failure due to biliary problems.
Patients with the DN type had the poorest outcome. Five patients (11.4%) were classified as this type and were followed up during a median (range) time of 11.5 (3.9–21.6) months. One patient (20.0%) died of biliary sepsis, and 2 (40.0%) underwent retransplantation. The remaining 2 patients (40.0%) did not recover from their life-threatening problems. Recurrent cholangitis, persistent jaundice, and poor nutritional status were a constant threat to their survival during follow-up. Prompt retransplantation, not repeated intervention and aggressive medical support, were required.
The results of this study showed that IHBS might progress to a more severe form by approximately 6 months after LT. Some patients had newly developed IHBS or progressed to a more severe form several months later. Figure 5 shows one example of a 54-year-old man. This patient had good patency of the biliary tree 1 month after LT; however, the biliary tree was substantially destroyed 4 months later. The patient underwent retransplantation 11.5 months after the primary LT.
The overall 2-year survival rate of patients with IHBS was 90.9%, excluding early deaths from other causes such as primary nonfunction and hepatic artery thrombosis, before returning to Korea. All patients in the UF or CO group were alive during the follow-up period. However, 2-year survival of the BM group was 85.7%, and there was no long-term survival expected in patients in the DN group (P = 0.024) (Fig. 6).
The incidence of IHBS without hepatic artery thrombosis after LT has been reported to be low (1.4%).14 Our study, however, showed that the incidence of IHBS may be as high as 25.6% in a particular group of patients who undergo LT from uncontrolled NHBD. IHBS might be one of the most important problems to overcome for the successful use of NHBDs for LT.
To assess prognosis, we divided the patients into good or poor prognosis groups and compared the prognosis for each classification type (Table 2). In cases where there was no requirement for intervention or where there was marked improvement of symptoms after the intervention, the prognosis was considered to be good; patients with persistent requirement for intervention without symptomatic improvement, retransplantation, or death due to biliary problems were considered to have a poor prognosis. Many patients classified as UF or CO did not require any intervention; even for patients who required several interventions, the result was great improvement of symptoms. According to our estimates, the UF or CO type stricture did not affect long-term survival despite possible symptomatic attacks. On the other hand, patients classified as BM or DN had poorer results. About half of the patients classified as BM were not responsive to repeated interventions; for the patients classified as DN, there was no recovery from life-threatening problems. All patients in the DN group have either undergone or require retransplantation.
No requirement of intervention; marked improvement of symptom after intervention.
Persistent need of intervention; retransplantation; death due to biliary problems.
Unilateral focal (n = 8)
Confluence (n = 10)
Bilateral multifocal (n = 21)
Diffuse necrosis (n = 5)
A management algorithm for post-LT IHBS that is based on our results is presented in Figure 7. Interventional therapy was implemented for patients who developed definite symptoms such as jaundice, itching, fever and chills, or hepatic dysfunction. Sixteen patients (36.4%) did not require any intervention despite the presence of definite strictures on cholangiography. Therefore, aggressive intervention may be unnecessary in asymptomatic patients.
Our study shows that patients classified as UF or CO were easily controlled with a few interventions. Aggressive repeated interventions were successful in 28.6% of patients classified as BM—a better result than previously reported.14 Thus, the first treatment attempt for IHBS should be interventional therapies. However, about half of the patients classified as having the BM type experienced no improvement in symptoms despite aggressive interventions. Therefore, retransplantation should be considered if symptomatic improvement does not occur by 1 year after transplantation, or if the patient's condition gradually deteriorates.
Finally, our results showed that the DN type classification was not controlled by interventional therapies. These findings are consistent with several previous reports of a high rate of graft failure.14, 17, 24 However, patient survival was not greatly affected if retransplantation occurred in a timely fashion.14, 15 Thus, retransplantation appears to be the most effective and perhaps the only way to manage the DN type. Although aggressive medical interventions should be performed first, retransplantation must be planned immediately after the diagnosis of a DN type of IHBS. The results of this study as well as previous reports show that biliary strictures may progress to a more severe form by approximately 6 months after LT.14, 25 Therefore, careful observation is needed even in patients with a mild form of IHBS with the expectation that problems may develop and additional strategies may be needed for treatment. The presence of a T tube helped us evaluate the biliary tree and to decide on initial intervention. Cholangiography via T tube is less expensive than magnetic resonance cholangiography, and it is less invasive than endoscopic or transhepatic cholangiography. In addition, many patients with mild IHBS have symptoms that will resolve only after intervention through the T-tube tract. Therefore, T-tube placement in the bile duct of patients at high risk for IHBS (e.g., those with NHBD LT, prolonged ischemic time, or ABO incompatibility) should be considered; removal can be considered at 6 months after LT.
In conclusion, IHBS occurred more frequently in uncontrolled NHBD LT than in conventional brain-dead DDLT or living donor LT; many of these patients were unlikely to be responsive to aggressive interventions. The UF and CO classification types had a good prognosis; patients in these groups experienced improvement in symptoms after several interventional treatments. However, in the absence of retransplantation, half of the BM patients and all of the DN patients had life-threatening problems despite aggressive, repeated intervention and medical support. Early retransplantation may be the only treatment option for these patients.