Liver transplantation in adults: Endoscopic management of biliary adverse events

INTRODUCTION — Liver transplantation (LT) is performed for patients with decompensated cirrhosis, hepatocellular carcinoma, and/or acute liver failure. Biliary tract adverse events including biliary strictures and bile leaks may contribute to patient morbidity and limited graft survival. Most biliary adverse events can be managed with interventional endoscopic retrograde cholangiopancreatography (ERCP) for LT recipients with duct-to-duct biliary anastomosis. Some LT recipients with Roux-en-Y anastomosis can also be managed endoscopically. (See "ERCP in patients with Roux-en-Y anatomy".)

This topic will discuss endoscopic management of biliary adverse events related to LT. Medical care of the LT recipient is discussed separately:

●(See "Liver transplantation in adults: Long-term management of transplant recipients".)

●(See "Liver transplantation in adults: Initial and maintenance immunosuppression".)

●(See "Infectious complications in liver transplantation".)

An overview of ERCP is discussed separately. (See "Overview of endoscopic retrograde cholangiopancreatography (ERCP) in adults".)

The prevention and management of infection related to ERCP (eg, cholangitis) are discussed separately. (See "Infectious adverse events related to endoscopic retrograde cholangiopancreatography (ERCP)".)

INCIDENCE AND RISK FACTORS — The estimated incidence of biliary adverse events related to LT ranges from 10 to 15 percent in recipients of deceased donor livers and from 15 to 30 percent in recipients of living donor livers [1-5]. Risk factors for biliary adverse events after LT include (table 1) [2,6-18]:

●Donor-related factors – Transplanted liver from a living donor (rather than deceased donor) or a donation after cardiac death donor; older age of donor; high donor body mass index; macrovesicular graft steatosis >25 percent [19-21] (see 'Special populations' below).

●Recipient-related factors – History of primary sclerosing cholangitis, cytomegalovirus infection.

●Vascular factors – Hepatic artery thrombosis, hepatic artery stenosis.

●Surgical/technical factors – Excessive dissection of periductal tissue during procurement of the liver graft, excessive use of electrocautery for control of biliary ductal bleeding, tension at the anastomosis, small bile duct diameter of the graft, mismatched size between donor and recipient bile ducts, ischemia/reperfusion injury, prolonged cold and/or warm ischemia times.

●Postoperative factors – Bile leak (via local inflammation and fibrosis) (see 'Bile leaks' below).

●Other factors – ABO incompatibility.

Whether the type of biliary reconstruction (duct-to-duct choledochocholedochostomy or Roux-en-Y choledochojejunostomy) increases the risk of biliary adverse events is uncertain [22-25]. However, duct-to-duct anastomosis facilitates endoscopic access to the biliary tree and preserves the sphincter of Oddi, which in theory avoids reflux of luminal contents into the bile duct [7]. (See "ERCP in patients with Roux-en-Y anatomy".)

Historically, T-tube placement was associated with biliary adverse events, but T-tubes are rarely used in most centers that perform liver transplantation [26-28].

EVALUATION

When to suspect a biliary adverse event — The clinical presentation of biliary adverse events varies among LT recipients. At one extreme, some patients have asymptomatic elevations in liver enzymes, primarily in a cholestatic pattern (disproportionate elevation of the alkaline phosphatase compared with the aminotransferases) but without abdominal pain, pruritus, or jaundice. (See "Approach to the patient with abnormal liver tests", section on 'Patterns of liver test abnormalities'.)

However, other patients may present with fever, right upper quadrant abdominal pain, and/or jaundice. Occasionally, other symptoms such as anorexia and pruritus are reported. For patients with a bile leak, exposure of the peritoneum to bile usually results in abdominal pain. However, some patients with bile leak do not have abdominal pain, possibly because of immunosuppression and hepatic denervation [6,7,25]. (See 'Bile leaks' below.)

The evaluation of other suspected hepatobiliary adverse events following liver transplantation (eg, acute T-cell mediated rejection) is discussed separately. (See "Liver transplantation in adults: Clinical manifestations and diagnosis of acute T-cell mediated (cellular) rejection of the liver allograft", section on 'Diagnostic evaluation' and "Liver transplantation in adults: Long-term management of transplant recipients", section on 'Hepatobiliary complications'.)

Initial testing — For patients suspected of having a biliary adverse event based on symptoms and/or elevated liver enzymes, evaluation begins with a transabdominal liver ultrasound with a Doppler study of the hepatic vessels. Further testing is informed by ultrasound findings:

●Biliary abnormality – If liver ultrasound demonstrates features of bile duct obstruction (eg, biliary ductal dilation) or a fluid collection suggestive of a bile leak, a cholangiogram is usually obtained to confirm the diagnosis, further define the adverse event, and perform an intervention if needed (eg, biliary stent placement) [1,6,7,18,25,29]. For patients with suspected biliary obstruction, the approach to care is outlined in the algorithm (algorithm 1).  

Selecting a method for obtaining a cholangiogram is determined by the type of biliary reconstruction, the likelihood of therapeutic intervention, and the available expertise. For patients with a suspected biliary adverse event based on symptoms, laboratory studies, and ultrasound findings (eg, dilated bile duct), ERCP or percutaneous transhepatic cholangiography (PTC) is typically performed. For patients with duct-to-duct anastomosis, we proceed with ERCP for diagnostic confirmation and therapeutic intervention [30,31]. We generally reserve PTC for patients in whom ERCP was unsuccessful and for some patients with a Roux-en-Y choledochojejunostomy. In high-volume centers with experienced advanced endoscopists, ERCP can be successfully performed in some patients with a Roux-en-Y choledochojejunostomy via small bowel enteroscopy [32-34]. Methods for performing ERCP in patients with Roux-en-Y anatomy are discussed separately. (See "ERCP in patients with Roux-en-Y anatomy".)

The evaluation of biliary adverse events in the LT recipient often includes performing ERCP that is simultaneously diagnostic and therapeutic, although in the nontransplant setting, ERCP is usually reserved for therapeutic indications. (See "Overview of endoscopic retrograde cholangiopancreatography (ERCP) in adults", section on 'Indications'.)

●Vascular abnormality – If a vascular abnormality such as hepatic artery stenosis or occlusion is suspected by Doppler study, a computed tomography scan with angiogram is obtained to confirm the diagnosis [35-37].

●No biliary or vascular abnormality – Transabdominal ultrasound may not be sufficiently sensitive to detect biliary obstruction (sensitivity of ultrasound, 38 to 66 percent) [10,38,39]. Thus, a normal appearing biliary tree should not preclude further evaluation when a biliary tract adverse event is suspected. For such patients, magnetic resonance cholangiopancreatography (MRCP) is obtained before proceeding with an invasive cholangiogram (ERCP or PTC) (image 1) [40].

MRCP is preferred for patients in whom ERCP and/or PTC cannot be performed. As examples, ERCP may be unsuccessful in patients with Roux-en-Y choledochojejunostomy and a long Roux limb, and PTC may not be feasible in patients without dilated biliary ducts. (See "Percutaneous transhepatic cholangiography in adults".)  

MRCP is a reliable, noninvasive technique for detecting biliary adverse events after LT, and its use is supported by society guidelines [31,40-42]. A meta-analysis of 20 studies found that MRCP had a sensitivity of 95 percent and a specificity of 90 percent for diagnosing post-LT biliary strictures [31]. In a study including 165 LT recipients with suspected biliary adverse event, performing ERCP when there was evidence of bile duct obstruction or bile leak on advanced imaging, laboratory tests, or liver biopsy was associated with higher rates of a high-yield endoscopic procedure (ie, one in which a clinically important intervention was needed) compared with diagnostic ERCP (ie, performing ERCP without preprocedure advanced imaging such as MRCP) (90 versus 66 percent) [41]. Rates of serious adverse events were not significantly different between the groups.

Subsequent testing — For patients with a suspected biliary adverse event but without evidence of biliary abnormality on transabdominal ultrasound and MRCP, liver biopsy is often performed to exclude acute T-cell mediated rejection (TCMR) or other causes of elevated liver enzymes (eg, recurrence of primary disease). It is important to exclude biliary obstruction with imaging (eg, ultrasound or MRCP) prior to performing liver biopsy because performing liver biopsy in the setting of biliary obstruction may increase the risk of biopsy-related events such as bile leak and peritonitis [43]. (See 'Initial testing' above and "Approach to liver biopsy", section on 'Complications'.)

The clinical features and diagnosis of acute TCMR in the LT recipient are discussed in more detail separately. (See "Liver transplantation in adults: Clinical manifestations and diagnosis of acute T-cell mediated (cellular) rejection of the liver allograft".)

BILIARY STRICTURES

Incidence and risk factors — Biliary strictures account for approximately 40 percent of all biliary adverse events after LT, and they have been classified as anastomotic or non-anastomotic strictures [7]. In a systematic review of 61 studies including over 14,000 liver transplantations, the overall incidence of biliary stricture formation was 13 percent (ie, 12 percent reported among deceased donor recipients and 19 percent among living donor recipients) [44].

Strictures that occur early after LT (eg, within two months after transplantation) are often related to technical problems during surgery, whereas nonearly strictures are mainly due to vascular insufficiency, ischemia, and problems with healing and fibrosis [7,39]. In addition, approximately 20 percent of postoperative bile leaks progress to biliary strictures [15,45]. (See 'Bile leaks' below.)

Anastomotic strictures

Clinical features — The majority of anastomotic biliary strictures (ABS) develop between 2 to 12 months after LT, but may occur at any time after LT [46]. ABS have been defined as being single, short, and located within 5 to 10 mm of the biliary anastomosis [47]. On cholangiogram, the stricture typically appears as a thin narrowing in the area of the biliary anastomosis (image 2A-C).

In some patients, a transient narrowing of the anastomosis may become evident within one to two months after LT due to postoperative edema and inflammation [25]. This postoperative narrowing is usually not a true stricture and does not require endoscopic intervention unless liver biochemistries are elevated and a biliary adverse event is suspected. (See 'When to suspect a biliary adverse event' above.)

Initial intervention

Goals — The goals of endoscopic intervention with ERCP are to increase the diameter of the ABS as demonstrated on cholangiography, to improve flow of bile through the anastomosis, and to resolve symptoms and laboratory abnormalities associated with cholestasis (eg, jaundice, cholangitis, elevated liver enzymes) [5,42,48].

General measures — We administer antibiotic prophylaxis to liver transplant recipients undergoing ERCP who are at risk for incomplete biliary drainage (eg, those with multiple biliary strictures and/or ischemic cholangiopathy) (table 2). General measures for patients undergoing ERCP, including indications for antibiotic prophylaxis, are discussed separately. (See "Overview of endoscopic retrograde cholangiopancreatography (ERCP) in adults" and "Antibiotic prophylaxis for gastrointestinal endoscopic procedures", section on 'Endoscopic retrograde cholangiopancreatography (ERCP)'.)

Frequency and number of endoscopic sessions — The total number of endoscopic sessions required for stricture resolution usually depends on timing of stricture development:

●Early stricture (within two months after LT) – Early strictures generally respond to one session of endoscopic balloon dilation and/or plastic stent placement. For most patients, the stricture will resolve within three months, and the anastomosis will remain patent without further intervention.

●Late stricture (≥2 months after LT) – Most patients with late ABS undergo ERCP with balloon dilation and temporary stenting (ie, 6 to 12 months) with a fully covered, self-expandable metal stent (SEMS) or with plastic biliary stents. Plastic stents are usually exchanged at three-month intervals to avoid increased risk of stent occlusion and bacterial cholangitis. Most patients who undergo ERCP with plastic stent placement require several endoscopic interventions (ie, generally ranging from three to five endoscopic procedures) to achieve stricture resolution [46,49-62]. In contrast, a fully covered SEMS remains in place for approximately six months until it is removed endoscopically [63-66]. (See 'Selecting a stent' below.)

Selecting a stent — The choice of biliary stent is informed by several factors including location and length of post-transplant ABS, efficacy, risk of stent dysfunction (eg, stent migration, stent occlusion), endoscopist preference, and stent availability. Specific features of biliary stents (ie, material, covering, shape, size) are discussed separately. (See "Endoscopic stenting for malignant biliary obstruction", section on 'Types of biliary stents'.)

●Distal (extrahepatic) ABS – For patients with distal (extrahepatic) ABS who require biliary drainage, we typically place a fully covered SEMS that is 10 mm in diameter and 60 or 80 mm in length. Fully covered SEMS provide effective drainage and require fewer endoscopic procedures than plastic stents, and studies have suggested a shorter time to stricture resolution with metal stenting [59,63-67]. In a systematic review of four trials including 205 patients with ABS after liver transplantation, there were no significant differences in rates of stricture resolution, stricture recurrence, and overall adverse events in patients treated with covered SEMS compared with plastic stents [67]. Metal stents were associated with shorter total duration of stenting (mean difference [MD] -105 days, 95% CI -202 to -8 days) and fewer endoscopic procedures (MD -1.86, 95% CI -3.12 to -0.6). In another systematic review of 10 studies including 200 patients with ABS, covered SEMS placement was associated with stricture resolution rates of 82 percent when the stenting duration was three months or longer [59]. In a subsequent study including 41 patients who were followed for five years after either removal of covered SEMS or observed stent migration, rates for remaining stent-free at five years were 49 percent overall and 61 percent among 31 patients with over four months of covered SEMS indwelling time [68]. In 28 patients with stricture resolution after SEMS removal or observed migration after a median of five months stent indwelling time, the rate of no stricture recurrence at five years was 72 percent. Serious adverse events were reported in 16 patients (39 percent), and the most commonly reported event was cholangitis.

●Proximal ABS – For patients with proximal ABS, we typically use plastic stents because covered SEMS may obstruct the right or left bile ducts. Endoscopic therapy with plastic stent placement has been associated with long-term stricture resolution for most patients with ABS [46,55,58,59,61,62,69,70]. In a systematic review of eight studies including 440 patients with ABS, endoscopic therapy with placement of multiple plastic stents for 12 months or longer was associated with stricture resolution rates of 97 percent, with a stricture recurrence rate of 2 percent after mean follow-up ranging from 11 to 40 months [49,59,70]. Most patients with stricture recurrence were successfully managed with endoscopic therapy with repeat plastic stenting. In a subsequent study including 56 patients with ABS, a course of endoscopic therapy with plastic stenting that required a median of four ERCPs was associated with a stricture resolution rate of 98 percent [69]. ERCP-related adverse events occurred in 3 of 56 patients (5 percent). After a median follow-up of nearly six years after stent removal, 3 of 50 patients (6 percent) had recurrent ABS. All three patients were successfully treated with endoscopic therapy and were asymptomatic after further median follow-up of nearly six years.

Strategies to reduce the number of stent exchanges have included a multi-stenting technique with stent exchange for selected indications (eg, biliary obstruction). In a study of 83 patients with ABS who had ERCP with stenting, a maximum of nine plastic stents were placed during the index procedure [70]. ERCP was repeated if biliary obstruction developed, if two or fewer stents remained, or if one year had passed without stent exchange. This approach was associated with a stricture resolution rate of 94 percent, while four patients (6 percent) required surgery (hepaticojejunostomy). After a median follow-up of 11 months, two patients (3 percent) required repeat ERCP for recurrent ABS.

Technique — The technique for stent placement is informed by stent type (see "Overview of endoscopic retrograde cholangiopancreatography (ERCP) in adults" and "Endoscopic biliary sphincterotomy"):

For fully covered SEMS placement:

●Place a guidewire across the stricture

●Place a covered SEMS across the biliary stricture

After the index procedure, ERCP is repeated in approximately six months to remove the covered SEMS and assess the stricture. Stricture resolution is defined by absence of the stricture on fluoroscopy and no resistance with a balloon (8 to 10 mm) sweep from the proximal to the distal bile duct [61]. If the stricture persists after stent removal, we place another fully covered SEMS across the stricture. After three to six months, we remove the covered SEMS endoscopically.

For plastic stent placement:

●Place a guidewire across the stricture

●Perform biliary sphincterotomy

●Dilate the stricture using balloons with diameters ranging from 6 to 8 mm

●Place one to three plastic stents (size range: 7 to 10 Fr) in a parallel fashion across the stricture

After the index procedure, ERCP is repeated in approximately three months for the following interventions: the previously-placed stents are removed, the stricture is examined and dilated if needed, and new stents are replaced across the stricture if they are still warranted. An increasing number of stents are generally used at each subsequent endoscopic session to achieve stricture resolution.

Subsequent intervention — For patients who do not respond to ERCP-guided stenting or who have recurrent ABS after responding to an initial course of endoscopic therapy, subsequent options include:

●Repeat course of endoscopic therapy – Repeat endoscopic therapy with stent placement has been associated with long-term success. In a study of 24 patients with recurrent ABS related to liver transplantation, 20 patients (83 percent) had long-term resolution of the stricture following at least one additional course of endoscopic therapy [71].

●Non-endoscopic interventions – Endoscopic therapy for ABS may be unsuccessful due to any of the following:

•For patients with fibrotic and/or angulated ABS, the guidewire rarely cannot be placed across the stricture during ERCP, a step that is necessary prior to stent placement [51,53,55,72,73]. Some patients require surgery for definitive therapy [74].

•For patients with Roux-en-Y choledochojejunostomy, ERCP may be unsuccessful, despite use of small bowel enteroscopy or other advanced endoscopic approaches. In such patients, percutaneous transhepatic cholangiography (PTC) and stricture dilation can be performed by interventional radiology, followed by placement of a percutaneous transhepatic catheter [75]. (See "ERCP in patients with Roux-en-Y anatomy".)

Surgical revision may ultimately be required in patients with strictures that are refractory to endoscopic and/or percutaneous treatment. A Roux-en-Y choledochojejunostomy is usually performed in patients with duct-to-duct anastomosis. For patients who already have a Roux-en-Y anastomosis, repositioning the bile duct of the graft to a well-vascularized area may be required.

Monitoring after endoscopic therapy — After stricture resolution, we monitor patients periodically with liver biochemistries (aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and total bilirubin) (eg, every one to three months). If liver enzymes are elevated over baseline values, magnetic resonance cholangiography is obtained for further evaluation. Patients with ABS require long-term surveillance and clinical follow-up since strictures may recur. Based on clinical observation and experience, risk factors for recurrence include initial presentation >6 months after LT and very tight strictures.

Nonanastomotic strictures

Incidence and etiology — Non-anastomotic biliary strictures (NABS) account for 10 to 25 percent of all stricture-related adverse events after LT, with a reported incidence ranging from 0.5 to 10 percent [22,24,29,50,76-78].  

NABS result mainly from hepatic artery thrombosis or other forms of ischemia. Less commonly, they can be due to recurrence of the underlying disease such as primary sclerosing cholangitis. Impaired blood supply may cause injury to the peribiliary glands and vascular plexus. Damage to these structures has been associated with the development of NABS after transplantation [79]. NABS are more common when donor organs have been retrieved after cardiac death. (See "Liver transplantation in adults: Deceased donor evaluation and selection", section on 'Donation after circulatory death'.)

Clinical features

●Timing – NABS usually develop within three to six months after LT, although they can present up to one year following surgery [50,77].

●Location and appearance – NABS appear as an irregularity or narrowing of the lumen of the intra- or extrahepatic duct, and NABS have been defined as being located at least 1 cm proximal to the anastomosis [47,80]. There may be multiple strictures involving the hilum of the liver and intrahepatic ducts, causing a cholangiographic appearance that resembles primary sclerosing cholangitis. Biliary sludge can accumulate proximal to the strictures, leading to the formation of casts [81]. This may predispose the patient to develop recurrent episodes of cholangitis. (See 'Bile duct casts' below.)

Management — Endoscopic therapy for extrahepatic NABS typically consists of ERCP and stricture dilation with a 4 to 6 mm balloon, followed by sphincterotomy and placement of plastic stents (size range: 10 to 11.5 Fr). Plastic stents are replaced every three months for approximately 6 to 12 months, and usually these strictures may require prolonged treatment including balloon dilatations and removals of casts and debris [50].

NABS are often challenging to treat because they may be multiple, diffuse, and/or involve small intrahepatic ducts that may be difficult to reach endoscopically [47]. Approximately 50 percent of patients have a long-term response to endoscopic therapy with stricture dilation and stent placement [1,50,81-84]. Some patients may require other interventions for treating NABS such as PTC or surgical revision (eg, conversion of duct-to-duct anastomosis to hepaticojejunostomy). However, up to 50 percent of patients experience progressive disease despite endoscopic therapy, leading to retransplantation or death [6,50,77,81]. Diffuse intrahepatic bile duct strictures have been associated with increased risk of poor graft survival and in many instances will require retransplantation for patients who are suitable candidates. (See "Liver transplantation in adults: Patient selection and pretransplantation evaluation".)

As an example, in a study including 81 patients with NABS, 28 patients (35 percent) underwent interventional treatment (ERCP, PTC, or surgery) [81]. Intervention was associated with improvement in liver enzymes; however, other outcomes were not reported for this subgroup of patients. Among 59 patients with NABS and radiologic follow up after a median of nearly two years, 28 patients (42 percent) had progressive biliary abnormalities on cholangiography [81]. Cirrhosis or advanced fibrosis developed in 23 patients (28 percent). Although graft survival was compromised (73 percent at five years), patient survival was not affected. Thirteen patients (16 percent) underwent liver retransplantation.

BILE LEAKS

Incidence and etiology — Bile leaks may originate from the anastomosis, the cystic duct remnant, or from the cut surface of the liver (for recipients of a living donor liver) (image 3). Bile leaks are often related to technical issues such as lack of perfusion from the hepatic artery and tension at the anastomosis [48]. In a systematic review of 61 studies, bile leak rates for recipients of deceased donor liver transplantation were 7.8 percent and for recipients of living donor liver transplantation (LDLT) were 9.5 percent [85].

Bile leak following liver transplant is a risk factor for developing biliary strictures and thus requires prompt intervention [9].

Clinical features and diagnosis — Most bile leaks develop within four weeks after LT and usually occur at the anastomotic site. Bile leaks are suspected in patients with abdominal pain, peritonitis, and/or jaundice.

The diagnosis of a bile leak can usually be confirmed with a transabdominal ultrasound that demonstrates a fluid collection. Magnetic resonance cholangiopancreatography (MRCP) can also detect a bile leak if transabdominal ultrasound is nondiagnostic. Hepatobiliary scintigraphy scanning, which has a sensitivity of approximately 50 percent and a specificity of approximately 80 percent, may also be useful for evaluating suspected bile leaks when other imaging studies are equivocal [86].

Management — The goal of endoscopic therapy is to eliminate the transpapillary pressure gradient, thereby permitting preferential transpapillary bile flow rather than extravasation at the site of the leak.

Endoscopic management of bile leaks is informed by the type of biliary anastomosis and severity of the bile leak (eg, size of fluid collection):

●Patients with duct-to-duct anastomosis – For recipients with duct-to-duct anastomosis and bile leak, ERCP-guided placement of a plastic, transpapillary biliary stent, with or without biliary sphincterotomy is performed because this approach is effective for resolving most bile leaks [1,87-89]. The duration of plastic stenting is approximately two months because of the potential risk of delayed healing related to immunosuppression [6]. Stent exchange is required only if recurrent biliary obstruction is suspected (eg, jaundice, elevated liver enzymes). After six to eight weeks of stenting, most leaks have healed, and the stent is removed endoscopically.

For some patients with small or low-grade leaks (ie, a leak that requires near complete intrahepatic filling to demonstrate contrast extravasation), some advanced endoscopists perform ERCP with biliary sphincterotomy alone rather than ERCP with stent placement. However, published data supporting this approach are lacking. In a study including 80 patients with bile leaks, ERCP with plastic stent placement (with or without sphincterotomy) was associated with higher rates of bile leak resolution compared with sphincterotomy alone (94 versus 58 percent) [90]. Endoscopic approaches to managing bile leaks in the nontransplant setting are discussed separately. (See "Endoscopic management of postcholecystectomy biliary complications", section on 'Bile leak'.)

Endoscopic therapy has been effective for managing bile leaks related to LT [1,6,49,87,88,90]. In a systematic review of 34 studies including 370 LT recipients with bile leak who were treated with endoscopic biliary stent placement, 278 patients (82 percent) had complete resolution of the bile leak [88].

●Patients with Roux-en-Y anatomy – For recipients with Roux-en-Y hepaticojejunostomy, anastomotic bile leaks are less common. However, for such patients with a bile leak, ERCP is often not feasible due to anatomic difficulties in reaching the biliary anastomosis. Management usually involves percutaneous internal to external drainage. Bile leaks in patients with Roux-en-Y anatomy may require surgical management if less invasive methods for diverting bile flow are not successful. (See "ERCP in patients with Roux-en-Y anatomy".)

●Patients with a T-tube – For patients undergoing LT, most centers no longer use T-tubes (ie, a T-shaped draining tube that is inserted into the common bile duct and exits through the abdominal wall to the skin surface in the right upper quadrant). (See "Surgical common bile duct exploration", section on 'T-tube placement'.)

However, for recipients with a T-tube, small anastomotic leaks can be diagnosed with a T-tube cholangiogram and managed by leaving the tube open to drainage without further intervention.

Following removal of a T-tube, some patients will develop a bile leak that results from delay in T-tube tract maturation due to immunosuppression. A bile leak should be suspected in patients who develop abdominal pain when the T-tube is removed.

ERCP (with or without sphincterotomy) with placement of biliary stent may be indicated for such patients, particularly in patients with abdominal pain. An alternative approach consisting of medical management with analgesics and observation for up to one week is also reasonable. If symptoms persist, management with ERCP and placement of transpapillary biliary stents is indicated [1,6,29,49,91]. Surgery or a percutaneous transhepatic approach is reserved for patients in whom ERCP is unsuccessful.

BILE DUCT FILLING DEFECTS — Filling defects in the bile duct after LT can be due to stones, sludge, blood clots, casts, and/or migrated stents; most defects are caused by stones [1]. Reported rates of bile duct filling defects have ranged from 2 to 10 percent [24,49,89,92]. The potential mechanisms related to the formation of stones, sludge, and casts include bile duct stricture and/or obstruction, prolonged warm and cold ischemia times, and bacterial infection [6,25,93,94].

Bile duct stones — Stones usually develop late (>1 year) after LT (image 4). As an example, in one series of 37 LT recipients with biliary stones, bile duct stones were identified at a median of 19 months after LT [49]. Following ERCP with stone extraction, the stone recurrence rate was 17 percent after median of six months. For most patients with biliary stones, one ERCP session with biliary sphincterotomy is effective for clearing the bile duct; however, for some patients, two or more endoscopic sessions may be required [49,92].

Endoscopic management of bile duct stones and sludge is similar to the nontransplant setting, and is discussed in more detail separately. (See "Choledocholithiasis: Clinical manifestations, diagnosis, and management", section on 'Subsequent evaluation and management' and "Endoscopic management of bile duct stones".)

Bile duct casts — Bile duct casts are usually related to ischemic injury (eg, hepatic artery thrombosis) or diffuse stricture formation in the hilum of the liver graft [95]. Clearance of casts may be difficult to achieve with endoscopic methods. In one series, combined endoscopic and percutaneous methods were able to successfully clear the casts in 60 percent of patients [95]. Various combinations of endoscopic methods including sphincterotomy, balloon and basket extraction, stent placement, and lithotripsy are often necessary, and many patients will ultimately require management with percutaneous intervention.

Patients with Roux-en-Y choledochojejunostomy and bile duct stones are typically managed with percutaneous drainage.

Biliary cast syndrome is an uncommon adverse event of LT that is characterized by biliary casts and debris causing biliary obstruction and cholangitis after LT. In two case series including a total of 1242 LT recipients, reported rates of biliary cast syndrome were 3 percent [96,97]. The etiology of biliary cast syndrome is not well understood. Associated risk factors include hepatic artery stenosis and biliary strictures. Biliary cast syndrome has been associated with limited graft survival. Several endoscopic interventions have been described with variable success, and some patients have undergone retransplantation [96].

OTHER ADVERSE EVENTS — Other adverse events of liver transplantation that involve the biliary tree include:

●Biloma – In patients with bile duct necrosis secondary to hepatic artery thrombosis, bilomas can occur due to bile duct perforation and extravasation of bile into the hepatic parenchyma or the abdominal cavity. Most bilomas occur in the perihepatic area, outside of the liver. If the biloma occurs in the hepatic parenchyma and communicates with the biliary tree, it may resolve spontaneously or, in some cases, may be managed with endoscopy and transpapillary stent placement. However, in this setting, the role of endoscopy is more diagnostic than therapeutic, as disrupted bile ducts may not completely communicate with the extrahepatic ducts.

Large bilomas not communicating with the extrahepatic bile ducts are managed with percutaneous drainage and antibiotics. Surgery is indicated when the bile leak cannot be effectively controlled with nonsurgical methods. (See "Repair of common bile duct injuries".)

●Sphincter of Oddi disorder – Sphincter of Oddi disorder (SOD) has been described in 1 to 7 percent of LT recipients [22,49,87,98,99]. The pathogenesis is unclear; one hypothesis is that denervation of the common bile duct in the ampullary region (secondary to surgical intervention) leads to the development of a hypertonic sphincter, causing dilated ducts and cholestasis [7,93]. Data have suggested that papillary stenosis and functional biliary sphincter disorders (as defined by Rome IV criteria) are uncommon and occur in 1 percent in LT recipients [100,101].

SOD is suspected in patients with cholestasis and a uniformly dilated bile duct without filling defects; abdominal pain is not usually present. In most studies, the diagnosis of SOD was based upon clinical suspicion and the response to biliary sphincterotomy. Other causes of cholestasis or elevated liver enzymes (eg, acute T-cell mediated rejection) should be excluded. (See "Liver transplantation in adults: Clinical manifestations and diagnosis of acute T-cell mediated (cellular) rejection of the liver allograft", section on 'Diagnostic evaluation'.)

Management of SOD in LT recipients is similar to the nontransplant setting (ie, biliary sphincterotomy), and this is discussed separately [101].

SPECIAL POPULATIONS

Recipients of living donor liver transplantation — Biliary adverse events after living donor liver transplantation (LDLT) may be related to surgical factors (multiple and/or small caliber ducts used for anastomosis, type of reconstruction) and non-surgical factors (older donor age) [102-106]. (See "Living donor liver transplantation in adults", section on 'Recipient outcomes'.)

For LDLT, many transplantation centers perform duct-to-duct anastomosis rather than Roux-en-Y choledochojejunostomy because duct-to-duct anastomosis has the potential advantage of a shorter operation time, no contamination by bowel contents, and a preserved sphincter of Oddi [20]. However, data have suggested that the risk of biliary stricture was lower with Roux-en-Y anatomy. In a meta-analysis of six studies including 1286 recipients of LDLT, 260 patients (20 percent) developed anastomotic biliary stricture and 118 patients (9 percent) had bile leak. Roux-en-Y choledochojejunostomy was associated with a lower risk of anastomotic biliary stricture (ABS) compared with duct-to-duct anastomosis (odds ratio 0.45, 95% CI 0.31–0.64) [107]. Rates of bile leak were not significantly different between groups.

Management of biliary strictures in recipients of living donor livers is similar to recipients of deceased donor livers and typically consists of ERC with balloon dilation and plastic stent placement for patients with duct-to-duct anastomosis. Reported rates of stricture resolution with this approach have ranged from 50 to 79 percent [58,108-110]. For patients who do not respond to endoscopic therapy or for those with Roux-en-Y anatomy, percutaneous transhepatic biliary drainage with stricture dilation is an alternative option [111]. As an example, in a study of 110 patients with LDLT and duct-to-duct anastomosis, 38 patients (35 percent) developed an anastomotic stricture, and ERCP was attempted as initial therapy in all patients [58]. Thirty-two patients (84 percent) were managed with endoscopic therapy alone, while six patients required a combination of percutaneous transhepatic cholangiography initially to cross the stricture, followed by endoscopic therapy. A median of four endoscopic sessions were required for stricture resolution. For most patients, strictures remained patent during an average follow up of 70 months, while eight patients (21 percent) had recurrent stricture after initial treatment success. All patients with stricture recurrence were successfully retreated with endoscopic therapy. (See 'Biliary strictures' above.)

Use of fully covered SEMS has been reported as an alternative for managing biliary adverse events in LDLT recipients. In a small cohort of 13 LDLT recipients with biliary strictures with median follow-up of 15 months, covered SEMS placement was associated with stricture resolution rate of 92 percent [112]. Although preliminary studies suggested that use of fully covered SEMS was safe in LDLT recipients, long-term data are lacking [113,114].

Living liver donors — Living liver donors may experience biliary adverse events following donor hepatectomy. Data from the Adult-to-Adult Living Donor Liver Transplantation Cohort consortium on liver donor outcomes included two studies with total of 1133 liver donors. The reported rate for overall biliary adverse events was 10 percent and for bile leak was 9 percent [115-117]. Biliary adverse events in liver donors are seen more often with right lobe donation than with left lobe donation, and the management is similar to that described for liver transplantation recipients. (See "Living donor liver transplantation in adults".)

ENDOSCOPY-RELATED ADVERSE EVENTS — The risks of advanced endoscopic interventions are generally regarded as similar for LT transplant recipients and nontransplant patients. For LT recipients, the estimated overall adverse event rates of up to 9 percent have been reported [1,118]. The most common adverse events are pancreatitis, cholangitis, and sphincterotomy-related bleeding. Adverse events of endoscopic retrograde cholangiopancreatography are discussed in more detail separately:

●(See "Post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis".)

●(See "Infectious adverse events related to endoscopic retrograde cholangiopancreatography (ERCP)".)

●(See "Post-endoscopic retrograde cholangiopancreatography (ERCP) bleeding".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Liver transplantation".)

SUMMARY AND RECOMMENDATIONS

●Incidence and risk factors – The estimated incidence of biliary adverse events related to LT ranges from 10 to 15 percent in recipients of deceased donor livers and from 15 to 30 percent in recipients of living donor livers. Risk factors for biliary adverse events are listed in the table (table 1). (See 'Incidence and risk factors' above.)

●Evaluation – Biliary adverse events after LT are suspected in recipients with elevations in liver enzymes, primarily in a cholestatic pattern (ie, disproportionate elevation of alkaline phosphatase compared with aminotransferases) with or without symptoms (eg, fever, right upper quadrant abdominal pain, jaundice). For such patients, the evaluation begins with a transabdominal liver ultrasound with a Doppler study, and further testing is informed by ultrasound findings. (See 'Evaluation' above.)

For patients with suspected biliary obstruction, the approach to care is outlined in the algorithm (algorithm 1).

●Anastomotic biliary strictures – For most patients with anastomotic biliary stricture (ABS) and duct-to-duct anastomosis, we suggest ERCP-guided intervention rather than percutaneous transhepatic cholangiography (PTC)-guided intervention or surgery (Grade 2C). An ERCP-guided approach with stent placement has been associated with long-term stricture resolution. (See 'Biliary strictures' above.)

PTC is reserved for patients in whom ERCP is unsuccessful or for some patients with Roux-en-Y choledochojejunostomy. (See "ERCP in patients with Roux-en-Y anatomy".)

For patients with a fully covered, self-expandable metal stent (SEMS), we typically remove the stent endoscopically in six months.

For patients with plastic stent(s), we perform stent exchange approximately every three months for a total stent indwelling time of 6 to 12 months. Most patients with plastic stenting will require a total of three to five endoscopic sessions.

After stricture resolution, we monitor patients with liver enzymes every one to three months and obtain imaging (eg, magnetic resonance cholangiogram [MRC]) when there is a rise in liver enzymes over baseline values.

●Non-anastomotic biliary strictures – For most patients with non-anastomotic biliary strictures (NABS) for which endoscopic therapy would be feasible, we suggest ERCP-guided balloon dilatation and plastic stent placement rather than PTC-guided or surgical intervention (Grade 2C). However, NABS are challenging to treat endoscopically, and some patients will require PTC-guided intervention, surgical revision, or retransplantation if progressive disease with diffuse biliary stricturing develops. (See 'Nonanastomotic strictures' above.)

●Bile leak – For most patients with duct-to-duct anastomosis complicated by a bile leak after LT, we suggest ERCP-guided intervention rather than PTC-guided intervention or surgery (Grade 2C). An ERCP-guided approach has been effective for resolving bile leaks.

We perform ERCP with placement of a transpapillary plastic stent, with or without biliary sphincterotomy. Stents generally remain in place for approximately two months to allow for healing of the ductal defect. ERCP is then repeated to confirm that the bile leak has resolved and to remove the stents. (See 'Bile leaks' above.)

●Endoscopy-related adverse events – The risks of endoscopy-related adverse events for LT recipients are similar to those reported for patients in the nontransplant setting. The most common adverse events are pancreatitis, cholangitis, and sphincterotomy-related bleeding:

•(See "Post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis".)

•(See "Infectious adverse events related to endoscopic retrograde cholangiopancreatography (ERCP)".)

•(See "Post-endoscopic retrograde cholangiopancreatography (ERCP) bleeding".)