Endoscopic management of biliary leaks: Where are we now?

Abstract

Biliary leaks can arise as a consequence of cholecystectomy, liver surgery, liver transplant, or, less frequently, trauma. Early identification and characterization of these leaks are crucial, as they can significantly enhance patient outcomes by reducing morbidity and mortality. Traditionally, surgical repair has been the standard treatment; however, advancements in endoscopic techniques and tools have established endoscopic retrograde cholangiopancreatography (ERCP) as the primary approach for managing these often-complicated cases. Interventions such as sphincterotomy, nasobiliary drainage, and stent placement aim to alleviate the pressure within the bile duct, facilitating depressurization and promoting leak healing. Alongside ERCP, endoscopic ultrasound is playing an increasingly vital role in addressing challenging cases. Ongoing improvements in endoscopic technologies and methodologies offer promising prospects, often minimizing the need for invasive surgical interventions. Nonetheless, the management of biliary leaks continues to pose significant challenges for clinicians. An optimal approach for patients experiencing bile leakage should be determined on a case-by-case basis and discussed within a multidisciplinary team involving radiologists, endoscopists, and surgeons. This comprehensive review aims to elucidate the role of endoscopy in the management of various types of biliary leaks, providing clinicians with practical insights to navigate this complex field.

Key Words: Biliary leak; Bile duct injury; Post-cholecystectomy; Liver transplant; Liver trauma; Endoscopic retrograde cholangiopancreatography; Sphincterotomy; Biliary stent

Core Tip: Bile leaks can occur as a complication of abdominal surgery or, less commonly, due to trauma. Timely and accurate diagnosis, along with appropriate treatment, is crucial for improving patient outcomes. With advancements in endoscopic technology and expertise, endoscopic retrograde cholangiopancreatography has emerged as the primary treatment approach, reserving surgical options for more complex or recurrent cases. However, the absence of specific guidelines means that treatment often relies on local resources and professional experience. This review seeks to clarify the role of endoscopy in managing various types of biliary leaks, offering clinicians valuable insights into navigating this challenging field.

INTRODUCTION

A biliary leak is characterized as the escape of bile from any part of the biliary tree, including the sectoral ducts, right and left hepatic ducts, common hepatic duct, cystic duct, and gallbladder[1]. These leaks can arise as a surgical complication, most commonly following procedures such as cholecystectomy, partial hepatic resection, and liver transplantation; however, they may also occur after any type of abdominal surgery. In addition to surgical factors, biliary leaks can result as a technical complication following endoscopic retrograde cholangiopancreatography (ERCP) or because of blunt or penetrating trauma[1,2]. Regardless of the cause, these leaks may necessitate surgical reintervention, prolong hospital stays, increase medical costs, and significantly affect patient morbidity and mortality[3].

The clinical presentation of a biliary leak is typically evident in patients with drains in place, as bile flows into the drain reservoir. However, in patients without drains, symptoms can be non-specific and may include jaundice, pruritus, fatigue, nausea, vomiting, abdominal pain and distension, as well as complications such as cholangitis, peritonitis, and sepsis. Additionally, unexplained leukocytosis, elevated C-reactive protein levels, and abnormal cholestatic liver tests may be observed[1].

Abdominal ultrasonography is often the initial diagnostic tool due to its availability, low cost, and minimally invasive nature, although its sensitivity is less than 60%[4]. Computed tomography provides a more detailed assessment of intra-abdominal organs and fluid collection, including biliomas and abscesses. Hepatobiliary scintigraphy, using agents such as iminodiacetic acid (HIDA), paraisopropyl HIDA, or diisopropyl HIDA, is a sensitive method for detecting suspected bile leaks, as it can demonstrate radiotracer leakage into the peritoneal cavity, confirming the presence of bile. However, it does not accurately localize the injury site and is best reserved for select cases due to its limited availability and higher costs[5].

Magnetic resonance imaging (MRI), along with magnetic resonance cholangiopancreatography, is widely used for imaging the biliary tree, playing a central role in evaluating these patients by precisely identifying the point of bile leakage. Nevertheless, cholangiography remains the gold standard for delineating the biliary tree, as it facilitates both diagnostic and therapeutic interventions. This can be performed either anterogradely via percutaneous transhepatic biliary drainage (PTBD) or retrogradely through ERCP[1]. Furthermore, endoscopic ultrasound (EUS) can identify complicating factors such as intraductal stones, provide access to the bile duct system, and allow for internal drainage of collections[6].

Although most biliary leaks are detected after considerable delay due to nonspecific presentation, early management is extremely important. Biliary leaks have previously been managed surgically, but more recently the less invasive endoscopic approach with ERCP is more prevalent[2]. Nevertheless, the treatment plan should be individualized to the patient and based on interdisciplinary discussions amongst endoscopists, interventional radiologists, and hepato-biliary surgeons.

TREATMENT OF BILE DUCT LEAKS

Biliary leaks can be managed using ERCP, PTBD, or surgical re-intervention. In recent years, ERCP and PTBD—sometimes utilized in conjunction for cases involving fluid collections—have emerged as the primary treatment options for most cases. Nevertheless, the choice of procedure is often determined by local expertise and leakage severity[7]. Bile leaks are classified as low grade (or minor) and high grade (or major) based on their identification during cholangiography: Minor leaks are identified after complete opacification of the intrahepatic biliary system, while major leaks are detected prior to opacification[8].

The sphincter of Oddi maintains moderately steady basal pressures ranging from 10 mmHg to 15 mmHg, with strong contractions capable of reaching up to 150 mmHg[9]. Regardless of the leak's etiology, the objective of endoscopic treatment is to reduce the transpapillary pressure gradient, facilitating fluid outflow to the duodenum, promoting healing of the upstream bile rupture, and controlling systemic sepsis. This can be accomplished through biliary sphincterotomy, biliary stenting, or a combination of both methods. Various types of stents may be utilized, including plastic straight or pigtail stents and fully covered metallic stents. Typically, plastic stents are removed or replaced after 2-3 months or sooner if occlusion is suspected, while fully covered metallic stents may remain in place for a longer duration, generally 6-9 months[9]. Recently, biodegradable biliary stents have shown promise in small studies for treating post-cholecystectomy leaks, eliminating the need for repeated endoscopic interventions for stent removal[10].

CLASSIFICATION OF BILE DUCT INJURY

Several classification systems have been developed to describe the extent and nature of bile duct injuries, ultimately guiding appropriate treatment and management (Supplementary Table 1).

The Strasberg classification is a widely utilized system that identifies bile duct leaks occurring during laparoscopic surgery based on injury location[11-13]. It is divided into five types, with type E incorporating the Bismuth-Corlette classification (Figure 1)[14,15]. However, this system does not consider vascular involvement. The Neuhaus classification modifies Strasberg’s parameters to distinguish between endoscopic and surgical treatment approaches, yet it similarly neglects vascular involvement and does not differentiate between lesions located at or above the hepatic duct bifurcation[16].

Figure 1 Strasberg classification[15]. A: Bile leak from cystic duct stump or minor biliary radical in gallbladder fossa; B: Occluded right posterior sectoral duct; C: Bile leak from divided right posterior sectoral duct; D: Bile leak from main bile duct without major tissue loss; E₁: Transected main bile duct with a stricture more than 2 cm from the hilus; E₂: Transected main bile duct with a stricture less than 2 cm from the hilus; E₃: Stricture of the hilus with right and left ducts in communication; E₄: Stricture of the hilus with separation of right and left ducts; E₅: Stricture of the main bile duct and the right posterior sectoral duct. Citation: Chun K. Recent classifications of the common bile duct injury. Korean J Hepatobiliary Pancreat Surg 2014; 18: 69-72. Copyright^© The author(s) 2014. Published by The Korean Association of Hepato-Biliary-Pancreatic Surgery (Supplementary material).

In 1995, McMahon et al[17] proposed a straightforward definition of major and minor bile duct injuries. The Stewart-Way classification categorizes bile duct injuries into four classes based on the mechanism of injury and origin of the lesion, without differentiating between lesions at or above the bifurcation of the hepatic duct (Figure 2)[15,18]. Conversely, the Amsterdam (or Bergman) classification identifies four types of ductal injuries occurring during laparoscopic cholecystectomy and considers the treatment approach. It is noteworthy that endoscopic intervention is successful in nearly all type A lesions, while type D lesions typically require surgical management[19].

Figure 2 Stewart-Way classification[15]. Citation: Chun K. Recent classifications of the common bile duct injury. Korean J Hepatobiliary Pancreat Surg 2014; 18: 69-72. Copyright^© The author(s) 2014. Published by The Korean Association of Hepato-Biliary-Pancreatic Surgery (Supplementary material).

The Siewert classification categorizes bile duct injuries based on the timing of detection and the type of injury, while also considering the presence of vascular lesions; however, it does not differentiate based on lesion location relative to the hepatic duct bifurcation[20]. In 2007, Bektas et al[21] introduced the Hannover classification, which delineates 21 patterns of iatrogenic bile duct injuries that occur during laparoscopic cholecystectomy. The advantage of this classification lies in its detailed categorization, which illustrates the association between the types of injury and surgical treatments employed (Figure 3)[15].

Figure 3 Hannover classification[15]. Citation: Chun K. Recent classifications of the common bile duct injury. Korean J Hepatobiliary Pancreat Surg 2014; 18: 69-72. Copyright^© The author(s) 2014. Published by The Korean Association of Hepato-Biliary-Pancreatic Surgery (Supplementary material).

Finally, the ATOM classification, developed in 2013 by the European Association for Endoscopic Surgery, represents a comprehensive framework that considers anatomy, time to detection, and the mechanism of injury, integrating various existing classifications into a cohesive system[22].

POST-CHOLECYSTECTOMY BILE LEAKS

Biliary leaks represent a significant complication following cholecystectomy, with reported incidence rates ranging from 0.3% to 0.9% of cases[23,24]. Notably, in the context of laparoscopic surgery, the incidence of biliary leaks has surpassed that seen with open cholecystectomy. Despite advancements in technology and surgical techniques, the trend in biliary leakage rates has not shown a decline[25].

The cystic duct is identified as the most frequent site of injury, followed by damage to the ducts located within the gallbladder bed, as well as the common bile duct, common hepatic duct, and T-tube tracts[8]. The predominant cause of such injuries is often attributed to the misidentification of the common bile duct, common hepatic duct, or aberrant ducts, particularly those on the right side. Additional contributors to these complications include inappropriate clipping of the cystic duct, accidental thermal injuries, and disruption of a bile duct that may directly enter the gallbladder bed[26].

Risk factors associated with an increased likelihood of biliary leaks during cholecystectomy can be classified into three categories: (1) Patient factors; (2) Local factors; and (3) Extrinsic factors[27]. Patient-related factors encompass advanced age, severe obesity, previous abdominal surgeries leading to adhesions, and underlying liver diseases such as cirrhosis. Local factors include severe inflammation and fibrosis related to acute cholecystitis, which complicate laparoscopic techniques and heighten the risk of incomplete gallbladder resection and ductal injuries. Furthermore, concurrent hemorrhage and various anatomical variations—such as a short cystic duct, a cystic duct that parallels the common bile duct, the presence of a hepatocystic duct, accessory cystic ducts, and the duct of Luschka—may threaten bile duct integrity[28]. Extrinsic factors are primarily linked to the surgeon’s experience and surgical equipment functionality[29]. Additionally, the presence of strictures or stones in the biliary tree post-cholecystectomy can elevate the pressure within the biliary system, thereby promoting the likelihood of a bile leak[3].

In terms of management, there has been a notable shift from surgical to less invasive procedural approaches, particularly ERCP, which serves both diagnostic and therapeutic purposes. ERCP facilitates the identification of the precise site of bile duct leakage and allows for assessment of severity through fluoroscopy with the infusion of a pressurized contrast medium combined with an occlusive balloon[30]. Therapeutic procedures during ERCP may include biliary sphincterotomy, biliary stenting—either alone or in conjunction with sphincterotomy—and nasobiliary drainage, with or without sphincterotomy[31]. Empirical evidence suggests that these strategies are both safe and effective in managing biliary leaks[32]. Nonetheless, the absence of specific guidelines for endoscopic therapy of bile duct leakage leaves the decision-making process largely dependent on the endoscopist’s clinical judgment and local resource availability.

In 2018, the European Society for Gastrointestinal Endoscopy recommended the endoscopic placement of plastic stents for managing bile duct leaks not attributed to transection of the common bile duct or common hepatic duct[33]. Furthermore, it suggested the use of self-expanding metal stents (SEMS) in cases of refractory bile leaks; however, the guidance did not specify the role of sphincterotomy, whether utilized independently or in conjunction with stenting.

In addition to assessing the effectiveness of various endoscopic techniques for treating biliary leaks, several ongoing discussions pertain to the optimal diameter of stents, the efficacy of bridging leaks with stents, the appropriate timing for endoscopic intervention, and the management of refractory leakage cases. These critical issues require meticulous evaluation and further research to enhance patient outcomes in biliary leak management. Biliary leaks represent a significant complication following cholecystectomy, with reported incidence rates ranging from 0.3% to 0.9% of cases. Notably, in the context of laparoscopic surgery, the incidence of biliary leaks has surpassed that seen with open cholecystectomy. Despite advancements in technology and surgical techniques, the trend in biliary leakage rates has not decreased.

ENDOSCOPIC TECHNIQUE

Different endoscopic techniques for treating bile leaks have been reported, including sphincterotomy alone, biliary stenting with or without sphincterotomy and the placement of a nasobiliary drainage. There are no specific guidelines on the use of endoscopic therapy for biliary leaks. However, date from existing literature indicate that stenting, with or without sphincterotomy, has better outcomes and provides faster leak resolution compared to sphincterotomy alone[33].

While refraining from biliary sphincterotomy may help avoid complications like bleeding and perforation, it is essential for eliminating retained stones[34]. Conversely, the benefit of conducting sphincterotomy alone is that it allows for completing the treatment in one session, thereby minimizing the risks associated with additional procedures and lowering costs. Therefore, it should be considered a viable option in cases of low-grade leaks where stone clearance is confirmed and a follow-up endoscopy for stent removal would pose significant risks[35].

Placement of a 7 French (Fr) nasobiliary drainage bridging the leak site is an option in patients with underlying coagulopathies, as it prevent sphincterotomy and endoscopic papillary balloon dilation. Moreover, a nasobiliary catheter may be helpful in patients where drainage is insufficient due to the presence of large stones or in those who are too ill to undergo the procedure with fluoroscopy. Furthermore, this procedure offers multiple advantages, including the ability to perform active decompression of the common bile duct by aspiration, providing a route for repeat cholangiography for confirming leak closure, avoiding a second ERCP session, and preventing duodeno-biliary reflux, which increases the likelihood of bile stone formation and cholangitis. However, patients often report discomfort and prolonged hospital stays[36]. Transpapillary biliary stenting with fully covered SEMS (FC-SEMS) serves as a second-line treatment for patients who do not respond to plastic stents or nasobiliary drainage, achieving nearly a 100% success rate[37-40].

The primary advantage of SEMS lies in their larger diameter compared to plastic stents, potentially leading to quicker leak resolutions. Additionally, FC-SEMS feature a membrane that can directly seal the leak and halt leakage immediately once bridging occurs. Nonetheless, they carry risks of post-ERCP pancreatitis and stent migration, and when inserted into a narrow common bile duct, they may lead to circumferential ischemia and subsequent biliary stricture[41].

The backflow of duodenal fluid into the bile duct may increase the risk of gallstone formation[42]. To mitigate these issues, some researchers have suggested positioning FC-SEMS above rather than across the papilla, thus eliminating the need for sphincterotomy. Although results from a limited number of patients are preliminary, they are promising[43,44]. The authors recommend reserving this approach for patients with high-grade leaks, are at high risk of bleeding, or who are unable to undergo multiple close ERCP sessions[43].

Plastic and metal stent have different patency time, with metal stents offering greater durability. Plastic stents need to be replaced every 3 months, while metal stents could be left in place for about 6-9 months.

To overcome the risk of stent obstruction and replacement, biodegradable stents have recently been applied to post-cholecystectomy leaks. The obvious advantage is to avoid additional procedures for stent removal. Siiki et al[10] conducted a study on 32 patients with low grade Strasberg grade A bile leaks; eight patients were treated with biodegradable stents and 24 patients were treated with a single plastic stent. The authors reported similar clinical outcomes between the two groups, although the biodegradable group required fewer endoscopic or surgical re-interventions and had lower overall drain output. Moreover, although the cost of the biodegradable stents was ten-fold higher, this group had a lower overall cost of treatment. To conclude, the use of biodegradable stents in treating bile leaks holds great promise. However, overall costs and application to leaks other than Strasberg type A need to be further investigated.

STENT DIAMETER

Stents introduced during endotherapy are grouped into small diameter stents (< 10 Fr) and large diameter stents (≥ 10 Fr). Only few studies have analyzed the impact of stent diameter on treatment outcomes for biliary leakage. Interestingly, some studies[8,45] reported a non-significant risk of failure in the large-diameter stent group[46-50]. Moreover, time to leak closure was not significantly different between the two groups (4 days in the small diameter and 4-5 days in the large diameter stents)[47-49].

BRIDGING STENTS

A meta-analysis published in 2019 revealed that stents traversing the site of bile leak —regardless of whether it was accompanied by sphincterotomy—demonstrated the highest likelihood of healing biliary leaks compared to stenting with short stents (with or without sphincterotomy)[2]. However, the results may be influenced by bias due to the inclusion of biliary leaks of varying causes and severities, the combination of plastic and metal stents in the analysis, and the absence of specific details regarding leak sites and stent lengths.

In practice, when the bile duct diameter proximal to the leak site exceeds that of the stent, positioning the stent above the leak does not significantly improve the likelihood of closure. Conversely, FC-SEMS, which are larger in diameter than plastic stents, can effectively bridge and seal the leak, thereby increasing the likelihood of closure. For peripheral leaks, FC-SEMS may also be positioned above the papilla[44]. However, this technique should be reserved for high-grade refractory leaks.

TIMING OF ERCP

In 2019, Abbas et al[3] published the only study that evaluated the timing of ERCP in managing biliary leaks following cholecystectomy and its impact on clinical outcomes. A prior study conducted by Adler et al[50] in 2017 also addressed this topic, focusing on patients who underwent ERCP for post-surgical leakage, 70% of which were post-cholecystectomy cases. Both studies categorized the timing of ERCP in relation to bile leaks into three groups: (1) Emergent (within the first 24 hours); (2) Urgent (on days 2 or 3); and (3) Expectant (beyond 3 days). Neither study observed a significant difference in the rate of adverse events post-ERCP. Adler et al[50] reported that the 90-day mortality rate was lower in the group undergoing ERCP on days 2 to 3 (2.9%) compared to those treated on day 1 (7%) or after day 2 (10%), although these findings did not achieve statistical significance. Both authors noted a U-shaped relationship between mortality and ERCP timing; however, Abbas et al[3] interpreted these results as indicative of severity bias. Patients presenting in critical clinical condition are more likely to require emergent ERCP or experience delays until stabilization, while those in stable condition tend to have the procedure scheduled for days 2 or 3.

Overall, existing literature makes it difficult to draw definitive conclusions about the optimal timing of ERCP. Nonetheless, interventions for biliary leaks should not be postponed if a patient's clinical situation deteriorates or if there is an increase in percutaneous drain output.

REFRACTORY LEAKS

The optimal approach for managing complex and refractory bile leaks following cholecystectomy has yet to be clearly established. In 2015, Canena et al[51] reported outcomes from a cohort of 16 patients with various types of post-cholecystectomy biliary leaks that did not resolve after endoscopic interventions involving biliary sphincterotomy and the placement of a 10 Fr transpapillary stent. All patients received repeated ERCP and multiple plastic stents (MPSs). Clinical success was achieved in 10 out of 16 patients (62.5%), with a median of three (range 2-4) plastic stents per patient. For the remaining six patients for whom MPS placement was unsuccessful, repeated ERCP was performed alongside FC-SEMS placement. The metal stents were removed after a median of 26 days, and leak closure was successfully achieved in 100% of cases by the end of treatment. There were no reports of complications related to the procedures or stent placements. Notably, high-grade biliary leaks were identified as a significant predictor of treatment failure, leading the authors to recommend FC-SEMS for such cases[51].

Prior studies have also indicated the efficacy of FC-SEMS in managing refractory bile leaks, albeit in small case series[37,52,53]. However, the occurrence of late adverse events, such as de novo choledocholithiasis and strictures at the proximal edge of the previous FC-SEMS site has been documented with prolonged stenting[52]. This suggests that early removal of stents, ideally 10-15 days after bile drainage cessation, is effective and not associated with adverse outcomes.

In 2019, Mutignani et al[54] conducted a retrospective study on 17 patients with Bergmann type A bile leaks refractory to previous endoscopic biliary sphincterotomy with biliary stent or nasobiliary catheter placement. All patients had a percutaneous or surgical drain placed adjacent to the origin of the fistula, which is common in bilioma cases. The group showed that withdrawing the drain from the fistulous site, along with extending the prior sphincterotomy if necessary, could promote leak closure. The rationale is to eliminate low atmospheric pressure at the leak site to restore a favorable pressure gradient that directs bile flow into the duodenum.

Therefore, before classifying a leak as refractory, it is important to consider alternative approaches and underlying factors. Indeed, it is paramount to reassess the radiological imaging to exclude anatomical variants that could have contributed to the leak persistence.

TRANSECTED BILE DUCTS

The complete transection of a major bile duct presents significant challenges for treatment via ERCP and often necessitates major surgical intervention, such as Roux-en-Y hepaticojejunostomy, typically following external biliary drainage. Nonsurgical repair of a transected bile duct can be accomplished through a combined percutaneous-endoscopic approach involving an interventional radiologist. In this method, the endoscopist approaches the distal section of the transected duct retrogradely, gaining access to the papilla, while the radiologist accesses the proximal segment anterogradely, utilizing a skin entry point and traversing an intrahepatic bile duct. The objective is to achieve an endoscopic-radiologic rendezvous, allowing for the placement of a stent to facilitate biliary reconstruction. However, the challenge lies in coordinating the guidewire from one access point with the retrieval basket from the other, ensuring they meet at the same level within a bilioma[55].

To address this challenge, a technique has been proposed that combines ERCP with EUS-guided cholangiography (CERES)[56]. The initial phase of CERES involves the creation of an EUS-guided hepatico-enterostomy (EUS-HE), which may include hepatico-gastrostomy for the left hemiliver and hepatico-duodenostomy for the right hemiliver. This EUS-HE procedure establishes an internal drainage system, thereby eliminating the need for PTBD and reducing the risk of fluid and electrolyte loss. However, this method is only viable when the intrahepatic ducts are dilated. EUS-HE can serve as a temporary measure prior to surgical repair, provide definitive biliary drainage, or act as the first step in a series of planned endoscopic treatments aimed at recanalization[57]. Recanalization can be performed anterogradely under fluoroscopic guidance through a mature hepatico-enterostomy, which generally requires at least one week following stent placement[56,58]. Additionally, both anterograde and retrograde cholangioscopy may be employed to achieve the rendezvous under direct visualization. Despite successful rendezvous, it is reported that the late stricture recurrence rate remains around 10%[55,56].

POST LIVER TRANSPLANTATION LEAKS

Liver transplant is the only available treatment for patients with end-stage liver disease. However, biliary complications following transplantation remain a major problem and are a significant cause of morbidity and mortality. Biliary complications include bilioma, bile duct obstruction due to stones, clots or casts, hemobilia, mucocele and sphincter of Oddi dysfunction. However, biliary strictures and leaks are the most common, occurring in up to 25% of patients[59].

Bile leakages arise more often from the anastomotic site but also from the cystic duct stump, the T-tube tract, the graft cut site, or an impaired accessory bile duct, possibly due to ischemic insults[60]. Seven out of ten biliary leaks occurred during the first month after transplantation and are usually found at the anastomotic site[61]. Conversely, late leaks (> 4 weeks) are typically associated with T-tube removal and are often associated with severe strictures[62]. Persistent biliary fistula after removal of the T-tube is common, as the required immunosuppression for transplantation can delay the formation of a fibrous tract around the T-tube. For this reason, many surgeons have abandoned the routine use of a T-tube to guide biliary restoration[63].

In the setting of immunosuppression, some patients with post-liver transplantation leakages may not present with symptoms. Leaks should be suspected whenever there is elevation of bilirubin, change in cyclosporine levels or bile in ascitic fluid[64].

Historically, biliary leaks have been thought to be more frequent in living donor liver transplantation (LDLT) rather than in deceased donor liver transplantation (DDLT). A recent meta-analysis performed on 25 studies reporting the incidence of post-liver transplant biliary complications that the incidence of biliary complications is almost doubled in LDLT compared to DDLT (24.4% and 13.1%, respectively). However, when studies were split into those published before and after 2014, studies published before 2014 showed higher odds of biliary complications for LDLT, while studies published after showed comparable odds of complications in LDLT and DDLT[65].

A recent meta-analysis focusing on the predictors of biliary leaks and strictures after LDLT found that multiple ducts for anastomosis increase bile leakage risk, while recipient age, duct-to-duct anastomosis and warm ischemia time do not affect leak occurrence. On the other hand, bile leaks were identified as the strongest predictors of post-LDLT biliary strictures[66]. To decrease leakage occurrence when two or more biliary orifices are closely located (< 3 mm), surgeons can perform a ductoplasty, which creates a common orifice resulting in a single duct-to-duct anastomosis[67]. However, ductoplasty increases the risk of stricture onset. These facts highlight the importance of a wide preoperative assessment of the bile anatomy of the graft with magnetic resonance cholangiography. Generally, duct-to-duct anastomosis is preferred because a shorter surgery time allows easy endoscopic access to the papilla and preserves the sphincter of Oddi, avoiding the reflux of intestinal content into the biliary system and reducing the risk of cholangitis[68]. However, in cases where multiple bile ducts require ductoplasty, surgeons may consider a left lobe graft or switch to hepaticojejunostomy[66].

Although some authors report the use of HIDA or MRI to diagnose biliary leaks after liver transplantation, endoscopic cholangiography has become the diagnostic method of choice, as it concomitantly provides treatment. Transhepatic cholangiography with subsequent interventional measures can be used in patients with Roux-en-Y hepaticojejunostomy when the endoscopic access to the afferent limb is not feasible, although it can be demanding when intrahepatic ducts are not dilated[1]. The combination of endoscopic sphincterotomy (with or without stent placement) and percutaneous catheter drainage can be performed in patients with concomitant abundant or infected bilioma[68].

With advances in diagnostic and therapeutic procedures, interventional endoscopy has become the most common method to treat biliary complications after liver transplant, as it is less invasive, safe, effective, and convenient for the patient[68]. Whether a leak occurs in an anastomotic or non-anastomotic site, the approach should be the same. Different endoscopic techniques have been successfully used, including the placement of nasobiliary drainage, sphincterotomy and stent placement. Again, the goal is to decompress the biliary tree to guide reconstruction. However, compared to post-cholecystectomy leaks, leaks occurring in post-transplant patients take longer to heal. This fact appears to be related to immunosuppressive agent use. In a retrospective study, Zimmerman et al[69] reported that the median time to tube, stent, and drain-free status after a biliary leak was 1 month longer among DDLT recipients (2.3 months) compared to LDLT recipients (1.3 months), although the difference was not statistically significant.

Despite improvements in technology and skills and high success rates, post-transplant leakage therapy is still challenging. Bile duct anastomoses are often small, tortuous, angulated or located next to the hilum, all factors that make the endoscopic approach difficult. Moreover, the outflow of medium contrast by the bile leakage during ERCP may obscure the anastomosis, precluding the passage of the guidewire[70].

NASOBILIARY DRAINAGE

A nasobiliary drainage endoscopically placed proximal to the leakage site can be successfully used to treat post liver-transplant leaks[71]. The advantage is that it allows frequent follow-up cholangiography to confirm leakage closure and there is no need for additional endoscopy for removal. However, beyond causing patient discomfort, it requires longer hospital stays and careful nursing, particularly in post-liver transplant patients with encephalopathy, where there is risk of tube dislodgement and stoma disjunction[46,72]. Another drawback to this approach is the diversion of bile from the intestine, which may diminish drug absorption.

SPHINCTEROTOMY AND/OR STENT PLACEMENT

One study reported that nasobiliary drainage combined with plastic stent placement is the best choice for treating post-transplant bile leaks because it can prevent conditions resulting from catheter dislocation and it may have prophylactic effects on upcoming bile duct strictures[46]. However, most centers use a biliary stent alone to flatten the difference in pressure between the gut and bile duct.

Plastic stents are most commonly used because of their lower costs and easy insertion. However, plastic stents have smaller diameters and a higher risk of occlusion compared to metal stents. Multiple side-by-side stent techniques have been used with good outcomes, although they require multiple endoscopic procedures and are associated with higher risks and costs. Metal stents with larger diameters were initially developed to provide larger stricture dilation and prolong patency. Moreover, they require fewer ERCP sessions, ultimately reducing hospital stays and costs[73]. Studies have shown that metal stents are effective in healing refractory biliary leakage. However, stents must be covered to avoid the embedding of the metal into the bile duct and to allow easy removal[39]. The drawback of the metal stents is their potential for migration, which can be overcome with the introduction of anti-migration systems or by placing the stent above the papilla[60,74,75].

Sphincterotomy is performed to facilitate stent insertion and to decrease pancreatitis risk due to the compression of the pancreatic orifice by the stent, especially in multiple biliary stenting and SEMS placement[76]. Moreover, sphincterotomy is useful in cases of concomitant stones, clots, and casts requiring extraction. However, sphincterotomy exposes the patient to the risk of bleeding and may induce duodenal content reflux into the bile duct with consequent cholangitis and stent occlusion. To overcome these drawbacks, some researchers have introduced a modified plastic stent above the papilla orifice without performing sphincterotomy. The stent does not have a distal flap but a thread attached to the distal end that falls into the duodenum to allow removal[77]. However, this stent has only been studied to treat post-liver transplant strictures.

In patients with post-transplant leakage, the stent should be left in place for approximately 2 months compared to the usual 4-6 weeks when a stent is placed post-cholecystectomy due to delayed leakage healing related to the use of immunosuppressive agents[60]. This may be why the use of biodegradable stents has not been described in this setting.

As with bile leaks, there is a high incidence of concurrent sludge or stones that increase the pressure in the biliary system, promoting and maintaining the leak. Therefore, bile duct clearance should always be performed after stent removal[3,78]. Sphincterotomy alone is infrequently used but can be reserved for low-grade leaks.

PATIENTS WITH ALTERED ANATOMY

In patients with Roux-en-Y reconstruction, endoscopic treatment is still feasible but is more challenging. A conventional side-viewing duodenoscope may not reach the bilio-enteric anastomosis depending on the length of the afferent limb[79]. Forward-viewing endoscopes, such as a cap-assisted pediatric colonoscope, single and double balloon-assisted endoscopes, and the more recent spiral endoscope, are more suitable for this task. The use of an overtube may help maintain endoscope position before the papilla[60]. Some authors favored a more invasive approach in this setting, including inserting the enteroscope through a previously performed percutaneous gastrostomy or jejunostomy[79,80].

WHEN ERCP FAILS

When conventional ERCP fails, as described for post-cholecystectomy leaks, it is possible to combine the percutaneous transhepatic and endoscopic transpapillary accesses. This so-called “rendezvous” technique has many variants, but all are possible only if the intrahepatic bile ducts are dilated enough to be punctured. The endoscopist and radiologist can work simultaneously in the same session or separately in two sessions, depending on local practice[81]. For biliary anastomotic disruption, the guidewire may be grasped into a bilioma[82].

Direct cholangioscopy with the use of the single operator cholangioscopy system (SpyGlass DS^™, Boston Scientific Corp., Natick, MA, United States) allows a more definitive assessment of biliary complications and can be used for therapeutic purposes aimed at definitive restitution of bile flow[83]. Direct cholangioscopy enables direct visualization, tissue acquisition, dilation of strictures, and removal of debris, casts and stones. A German group studied the application of methylene blue under peroral cholangioscopy to detect irregularities and necrotic areas where conventional cholangioscopy showed few or no abnormalities. The dye is taken up by viable cells, while necrotic cells do not show any specific coloring. With this technique, extensive bile duct necroses that lead to bile leakage can be detected[84].

BILE LEAKAGES IN LIVING DONORS

Living donors are at risk for biliary complications. The most common complication is bile leakage, occurring in 9.2% of cases, primarily in right lobe transplants[85,86]. Leaks occur early in the postoperative phase, with a median of 14 days after surgery. However, they tend to resolve faster compared to transplant recipients[87]. ERCP is the gold standard for bile leak diagnosis and treatment, with similar approaches to those described for liver recipients.

POST-HEPATIC RESECTION BILE LEAKS

Progress in surgical techniques, better understanding of hepatic physiology and anatomy, and increasing experience in perioperative management have led to remarkable improvements in the hepatic surgery patient outcomes. Concurrently, the indications for liver resection have expanded and the complexity of liver resection has increased. More aggressive hepatectomy such as two-stage resection and Associating Liver Partition and Portal vein ligation for Staged procedures have been introduced in patients with benign and malignant diseases[88,89]. In this setting, the incidence of post-hepatectomy leaks has become one of the most important problems in liver surgery, ranging from about 3% to 20% and increasing up to 43% when combined with biliary reconstruction[90,91]. The occurrence of biliary leaks prolong hospital stay, delays the removal of abdominal drains, and increases costs and mortality rates.

Most bile leaks present early after liver surgery, generally on day 2-4, with bile flowing by percutaneous drains or spilling from surgical incisions. In patients without drains or with leakage originating far from the drain, leaks present more insidiously and later, generally on day 6-8 after surgery[92]. The International Study Group of Liver Surgery defines bile leaks as either 1) a bilirubin concentration in the drainage fluid that is at least three times higher than the serum bilirubin level, measured on or after postoperative day 3, or 2) the need for radiologic or surgical intervention due to biliary collections or bile peritonitis[93]. Bile leakages after hepatic surgery are classified as grade A when there are no or minimal consequences and no change to the patient’s clinical management is required; grade B leakages require radiological or endoscopic therapy, while grade C bile leaks require surgical re-intervention[93].

Bile leaks after hepatectomy are related to anatomical features than to patient comorbidities, with the type of resection and exposed parenchymal volume being those mostly responsible. Left side hepatectomy, hepatectomies exposing the major Glisson’s sheath or including the hepatic hilum and cholangiocarcinoma have been identified as risk factors for bile leaks[1]. In addition, cirrhosis underlies a state of immunodeficiency with reduced response to stimuli, combined with a persistent activation of the immune system cells producing pro-inflammatory cytokines[94].

ENDOSCOPIC THERAPY

While the endoscopic treatment of bile leaks after cholecystectomy or at the anastomotic side after liver transplantation is better established, the role of endoscopy to seal grade B biliary leaks complicating liver resection is less defined. Nevertheless, the endoscopic approaches resemble those previously described, such as sphincterotomy alone, stenting alone and sphincterotomy in combination with stenting (with or without bridging)[8,32]. Data on therapeutic success of these methods are limited. Schaible et al[95] reported an overall success treatment of 74%, in line with those stated by other authors[92,96-98], even though the type of endoscopic technique, leak severity and extent of surgical intervention differ among studies. However, higher rates of sustained therapeutic success were found in patients receiving bridging stents compared to those treated with stenting distal to the leak or sphincterotomy alone[95,98].

The ability of placing a bridging stent depends on the leak location. Leaks at the right hepatic duct, left hepatic duct or the hilum are easier to cannulate because they occur in larger caliber bile ducts closer to the papilla. On the other hand, peripheral leaks belonging to small caliber ducts further from the papilla are harder to reach by the guidewire. The rendezvous technique via a percutaneous or EUS-approach is still a valuable option when traditional cannulation fails or when anatomical variants occur.

BILE LEAKS FROM ISOLATED DUCTS

Bile leaks originating from isolated ducts, similarly to post-cholecystectomy Strasberg type C lesions, have not been approachable via endoscopy due to lack of connection between the leak and biliary tree. Mutignani et al[99] described an innovative method to take endoscopy out of the lumen to restore previously inaccessible bile leaks. After bile duct cannulation, cholangiography, sphincterotomy, and identification of the stump of the resected hepatic duct, a hydrophilic wire was advanced within a sphincterotome to the stump. When the sphincterotome was facing the stump, the hydrophilic guidewire was replaced with a stiff guidewire backwards into the sphincterotome. The hard tip of the guidewire was forcefully pushed to perforate the stump. Therefore, access to the peritoneal cavity was enlarged with a balloon dilation or a cysto-enterostome and the guidewire was changed in favor of a hydrophilic one. Hence, the soft tip of the hydrophilic wire was used to cannulate the isolated duct. When successful, a bridging transpapillary plastic or metal stent was placed, whereas a transpapillary stent was placed with the proximal edge into the peritoneal cavity in unsuccessful cases to drain the bilioma. The authors treated 13 patients: two patients with bridging stents and 11 patients with non-bridging stents. They achieved technical and clinical success in all cases. Therefore, bridging the duct is not mandatory for success. It is important to place the stent as close to the leak as possible. Moreover, the authors remarked on the importance of pulling back the abdominal drain from the fistula’s site, as the drain modifies the intra-abdominal pressure and could interfere with leak closure. Finally, regarding the choice between plastic and metal stents, the authors suggest the insertion of metal stents in case of cavities to allow better drainage of secretions and necrotic tissue. In the absence of a cavity, plastic stents should be sufficient[99].

ENDOSCOPIC INTERNAL DRAINAGE OF COMPLEX BILIOMAS

Nine out of ten biliomas resolve after the leak is treated with conventional endoscopy consisting of biliary sphincterotomy and/or stent placement[100]. However, in cases of high outflow biliary leaks, hilar or intrahepatic leaks, excluded ruptured hepatic segments or altered anatomy, conventional endoscopy should not be excluded if feasible, especially when considering additional surgical or radiologic intervention. These procedures expose the patient to additional risks and increase mortality rate, especially when hepatectomy is performed for oncological purposes[50,101]. A Belgian group proposed two different endoscopic techniques in addition to conventional endoscopy for managing biliomas secondary to refractory biliary leak: (1) Transpapillary/transfistulary drainage; and (2) EUS-guided transmural drainage[102]. Transpapillary/transfistulary drainage of bilioma involves the placement of a double-pigtail plastic stent (DPT) inserted through the papilla with the first pigtail inside the bilioma and the second pigtail in the duodenum. In cases where there is a long distance between the fistula and the papilla, two DPTs can be placed.

On the other hand, EUS-guided transmural drainage of bilioma consists of placing one or two DPTs through a transmural access to the bilioma from the stomach or the proximal duodenum. Authors reported a clinical success rate of 81.5% for both techniques with percutaneous drain removal and significant bilioma regression, with sharply lower need for surgery and long-term percutaneous drainage. The rate of serious adverse events related to the procedure was considerable (13%), mostly involving bleeding, with two deaths[102]. However, repeat surgery for biliary leaks is equally associated with a significant increase in mortality rate[103]. EUS-guided transmural drainage resulted in lower clinical success rates compared to transpapillary/transfistulary drainage. Therefore, the authors conclude that EUS-guided transmural drainage should be prioritized, besides being the only possible approach in case of altered anatomy. Nevertheless, the use of 7 Fr stents for their soft consistency and the recourse to abdominal CT to rule out close contact between stents and hepatic vessels are always suggested to decrease bleeding[102].

MEASURES TO PREVENT BILE LEAKS

Leakage points on the liver resection margins generally emerge with bile stain that are subsequently sutured. Therefore, small leaks are challenging to find. In this setting, many researchers have developed approaches to prevent the formation of hidden biliary leaks. Leak tests involve the injection of a substance through the biliary tree, commonly the remaining cystic duct, increasing the pressure of the bile duct and inducing the outflow of the substance from the plane of hepatic resection in case of rupture[104]. Many substances have been tested in addition to cholangiography, such as methylene blue dye, indocyanine green (ICG) dye, fat emulsion (white test), propofol, air, normal saline, topical hydrogen peroxide to the cut liver surface and narrow-band imaging approaches[94]. However, the impact of bile leak testing on postoperative bile leak rate and the most effective test substance to use are controversial. In addition, in patients where liver resection results in the exclusion of a bile segment and leaky bile ducts are not in communication with the biliary tree, bile leak testing cannot identify leak sites.

A recent meta-analysis showed that routine intraoperative bile leak tests significantly reduce the incidence of postoperative bile leakages, overall morbidity, need for reintervention, and duration of hospital stay[94]. However, the most effective substance and volume to use remain unclear. Adverse events, although rare, need to be considered. There has been evidence of allergic reactions and barotrauma during dye injection; for white tests in particular, there is a risk of fat embolization. Normal saline and fat emulsion are easily washed out compared to methylene blue and ICG, so their sequential use is suggested[94].

Another method to prevent bile leaks after hepatic resection has been the use of fibrin glue to coat the cut liver edge. However, there is not strong evidence that fibrin sealants reduce the incidence of bile leakage after liver resection. In vitro, bile contains profibrinolytic activity that causes lysis of the clot formed by the fibrin sealant. Therefore, although fibrin sealants can be useful to achieve hemostasis, their routine use in liver surgery is not recommended[105].

Finally, the use of omentum to cover the cut edge of the liver has been tested with the rationale of delivering blood-borne hemostatic agents and leukocytes through its net of capillaries and lymphatics. However, this method did not decrease the incidence of bile leaks, with a drawback of requiring postoperative percutaneous drainage[106].

BLUNT AND PENETRATING TRAUMA TO THE BILIARY TRACT

In the hepatic trauma field, biliary injury occurs in 0.5%-21% of cases, depending on the criteria and methods used to diagnose bile leakages[107]. Approximately 80%-90% of biliary injuries occur after sharp penetrating traumas like stab or gunshots, while blunt traumas are less common (2.8%-7.4%) and are caused mainly by road traffic accidents, falls, or kicks[108].

Biliary injuries can be intrahepatic or extrahepatic, involving the gallbladder (especially when distended) or a combination. Traumatic biliary tract injuries are often associated with injuries of other intrabdominal organs such as spleen, duodenum and pancreas, making diagnosis and treatment even more challenging than that of iatrogenic bile lesions[109]. Moreover, vascular complications requiring embolization or even surgery, especially in case of large devascularization, may concomitantly occur, complicating diagnosis and therapy[110]. It is unclear whether bile leak location is prognostic of treatment outcome[111]. Conversely, the time of diagnosis and intervention method play a significant role in decreasing morbidity and mortality[112]. Biliary complications such as bilioma can present even weeks after initial trauma, often with nonspecific progressive symptoms[107].

In 1994, the American Association for the Surgery of Trauma introduced a classification system for grading liver injury severity to facilitate clinical investigation and outcomes. This classification can be used to predict patients at risk for bile leaks (Table 1)[113].

Table 1 Liver injury scale (1994 revision).

Grade (advance one grade for multiple injuries, up to grade III)	Injury description
I	Hematoma: Subcapsular, < 10% surface area
	Laceration: Capsular tear, < 1 cm parenchymal depth
II	Hematoma: Subcapsular, 10%-50% surface area; intraparenchymal, < 10 cm in diameter
	Laceration: 1-3 cm parenchymal depth, < 10 cm in length
III	Hematoma: Subcapsular, > 50% surface area or expanding; ruptured subcapsular or parenchymal hematoma intraparenchymal hematoma > 10 cm or expanding
	Laceration: > 3 cm parenchymal depth
IV	Laceration: Parenchymal disruption involving 25%-75% of hepatic lobe or 1-3 Couinaud’s segments within a single lobe
V	Laceration: Parenchymal disruption involving > 75% of hepatic lobe or > 3 Couinaud’s segments within a single lobe

In recent years, the treatment of hemodynamically stable patients with sustained hepatic trauma has shifted from definitive surgical treatment to nonoperative management. Again, ERCP has both diagnostic and therapeutic abilities. Sphincterotomy with/without stent placement can successfully manage many complications and has sharply decreased mortality rate[110]. However, location and extent of biliary disruption may hamper the achievement of the goal. In these cases, PTBD placement can provide decompression and/or diversion of bile leaks. Nevertheless, the decision relies on the extent of the biliary injury, association with other organ injuries and local expertise[114].

In conclusion, management of traumatic bile injuries remains a challenging dilemma and requires multi-modal therapy. ERCP can avoid unnecessary invasive procedures, although percutaneous and surgical procedures may be required in severe cases. In all patients, treatment must be individually tailored in a multidisciplinary approach.

CONCLUSION

A bile leak can occur from a variety of sources including iatrogenic and traumatic. Prompt identification and treatment are extremely important to decrease morbidity and mortality and facilitate optical outcomes. At present, ERCP is the standard of care. All endoscopic techniques aim at decreasing pressure inside the bile duct by promoting decompression with sphincterotomy or stent placement. In recent years, innovations in endoscopic techniques have substantially improved, so that surgical measures are reserved for serious or recurrent cases. In the future, more developments involving ERCP and EUS are still expected. Nevertheless, biliary leaks are still challenging for clinicians, and the most beneficial treatment option must be chosen using a multidisciplinary approach.