Biliary drainage in patients with altered anatomy: Literature review of different endoscopic approaches

Abstract

Endoscopic retrograde cholangiopancreatography (ERCP) in patients with surgically altered anatomy remains a challenging field in therapeutic endoscopy due to the complex anatomical reconstructions that limit access to the biliary tree. Over the past two decades, device-assisted enteroscopy (DAE), including single-balloon, double-balloon, and motorized spiral enteroscopy, has expanded the feasibility of ERCP in this population, with overall technical success rates generally reported between 70% and 90%. Nevertheless, these techniques are technically demanding, time-consuming, and frequently affected by limited reach and unstable positioning. More recently, interventional endoscopic ultrasound (EUS)-guided procedures have emerged as highly effective alternatives, significantly improving clinical outcomes in selected patients, particularly in those with long-limb Roux-en-Y reconstructions where conventional methods are less effective. Percutaneous transhepatic biliary drainage continues to represent a valuable salvage option when endoscopic approaches fail, though it is associated with a greater burden of reinterventions and adverse events. This minireview provides a comprehensive overview of the main endoscopic strategies for biliary drainage in altered anatomy, focusing on technical considerations, efficacy, and safety profiles of DAE-assisted ERCP, EUS-guided interventions, and motorized systems. The evolving landscape of biliary drainage in this setting highlights the need for tailored treatment strategies, multidisciplinary collaboration, referral to high-volume centers, and further prospective studies to refine patient selection and optimize clinical outcomes.

Key Words: Surgical altered anatomy; Enteroscopy; Endoscopic ultrasound; Biliary drainage; Endoscopic retrograde cholangiopancreatography

Core Tip: In patients with altered anatomy, endoscopic retrograde cholangiopancreatography is technically demanding and requires expertise in device-assisted enteroscopy. Success depends on anticipating limitations such as scope instability, restricted reach, and prolonged procedure time, and on recognizing when to switch to alternative strategies. Endoscopic ultrasound-guided drainage has become an important option in long-limb reconstructions, while percutaneous transhepatic biliary drainage remains the most frequently used rescue technique, with surgical alternatives still possible in selected cases. Highlighting these practical considerations helps distinguish expert-level decision-making from a general overview.

INTRODUCTION

Biliary drainage (BD) in patients with altered gastrointestinal (GI) anatomy represents a significant challenge for therapeutic endoscopy. Surgical modifications, such as Roux-en-Y gastric bypass (RYGB), Billroth II gastrectomy, and hepaticojejunostomy, often render conventional endoscopic retrograde cholangiopancreatography (ERCP) unfeasible. To address these anatomical complexities, several alternative strategies have been developed and refined over recent years.

Techniques such as enteroscopy-assisted ERCP (using single-balloon or double-balloon enteroscopes), endoscopic ultrasound-guided BD (EUS-BD), and percutaneous transhepatic BD (PTBD) have emerged as valuable approaches. Each method offers distinct advantages and limitations depending on the clinical scenario, anatomical alterations, and available expertise.

The number of requests for BD in patients with a surgically altered anatomy (SAA) is increasing, due to the growing incidence of bariatric surgery and improved survival rates among patients with surgically treated upper GI cancers[1,2].

This review aims to provide a comprehensive analysis of the existing literature on endoscopic BD techniques in altered anatomy. By comparing the efficacy, safety, and technical considerations of these approaches, we seek to offer insights to optimize patient management and identify potential areas for future innovation.

TYPES OF SURGICAL PROCEDURES RESULTING IN ALTERED ANATOMY

Various surgical interventions significantly modify the GI tract’s anatomy, presenting unique challenges for endoscopic BD and necessitating specialized management approaches.

RYGB - Roux and Y reconstruction after distal or total gastrectomy

Gastric bypass creates a small gastric pouch (15-30 mL) separated from the remnant stomach, excluding the duodenum, pylorus, and antrum from the alimentary stream. A Roux-en-Y jejunojejunostomy reconnects the alimentary and biliopancreatic limbs 75-150 cm from the ligament of Treitz (Figure 1). This reconstruction, also used after gastric cancer resections, reduces biliary reflux and remnant gastritis compared to Billroth II[3]. Minimally invasive approaches often use a double-loop technique with sequential gastro- or esophagojejunostomy and jejunojejunostomy, divided by a linear stapler[4].

Figure 1  Roux-en-Y gastric bypass: Endoscopic view of anastomosis.

Billroth II reconstruction after distal gastrectomy

This reconstruction involves a distal gastrectomy with a gastrojejunostomy, where the gastric remnant is anastomosed to a jejunal loop approximately 20-30 cm distal to the ligament of Treitz, either through an antecolic or retrocolic route. The duodenal stump is closed, excluding the duodenum from the alimentary passage. To reduce bile reflux and prevent afferent loop syndrome, a Braun enteroenterostomy may be created between the afferent and efferent limbs[3,5].

Whipple procedure (pancreaticoduodenectomy)

This operation entails removal of the pancreatic head, duodenum, gallbladder, distal common bile duct, and often the gastric antrum. Reconstruction is performed with a pancreaticojejunostomy, hepaticojejunostomy, and gastrojejunostomy (or esophagojejunostomy if the stomach is resected). The ampulla of Vater and duodenum are entirely removed, and biliary and pancreatic secretions are diverted into the jejunum through separate anastomoses[6].

Biliary-enteric anastomoses (e.g., hepaticojejunostomy, choledochojejunostomy)

These procedures are performed for benign strictures, bile duct injuries, or malignant obstruction. The proximal bile duct (common hepatic or common bile duct) is anastomosed to a Roux-en-Y jejunal limb 40-60 cm in length. As the bile duct drains directly into the jejunum, the duodenum and papilla are bypassed, and the long limb with absent landmarks makes endoscopic access particularly challenging.

Biliopancreatic diversion with duodenal switch

This bariatric procedure includes: (1) Sleeve gastrectomy or subtotal gastrectomy; (2) Transection of the duodenum just distal to the pylorus; (3) Anastomosis of the proximal duodenum to the distal ileum (alimentary limb); and (4) The bile and pancreatic ducts drain into the long biliary limb, which joins the alimentary limb approximately 50-100 cm proximal to the ileocecal valve[7,8].

ENDOSCOPIC BD OF SAA PATIENTS

Enteroscopy-assisted ERCP

Surgically altered GI anatomy is a condition that may predispose to the development of biliary adverse events (AEs) like cholangitis due to bilioenteric anastomotic stenosis or bile duct stones[2]. The wide variety of biliary tract pathologies, such as stones and benign or malignant biliary strictures, can lead to hepatobiliary dysfunction, cholangitis, and ultimately, liver failure, making adequate therapy imperative. In this context, ERCP is the gold standard for the diagnosis and treatment of these diseases[9,10]. However, in patients with SAA, ERCP with standard side-viewing endoscopes is often difficult or unfeasible due to the length and tortuosity of the surgical segments[11]. Enteroscopy-assisted ERCP in altered anatomy patients is a technically complex procedure due to several factors, including the limited maneuverability of the endoscope caused by flotation, looping, and adhesions of the small bowel (SB), difficulty in recognizing the afferent limb and the distance between the anastomosis and the papilla that often is observed in the opposite view from the normal anatomy. The scope choise is a central determinant of success in ERCP for patients with SAA, and should be tailored to the specific surgical reconstruction. In Billroth II gastrectomy, where the afferent loop is typically short, a forward-viewing endoscope such as an operative gastroscope or a pediatric colonoscope can be sufficient. Pediatric colonoscopes, with their longer working length (133-168 cm), slim outer diameter, and relatively large working channel (3.2 mm in some models), provide advantages in these cases and can avoid the need for balloon-assisted enteroscopy (BAE). A large retrospective study found ERCP with forward-viewing endoscopes to be safe and effective for a variety of SAA types[12]. Their utility has also been demonstrated in short-limb Roux-en-Y reconstructions, particularly when the papilla or bilioenteric anastomosis is not too distant from the gastrojejunostomy. Wang et al[5] reported that ERCP performed with a pediatric colonoscope in patients with Roux-en-Y gastrectomy and intact papilla achieved high technical and clinical success rates with acceptable safety, suggesting that this strategy may represent a first-line option when BAE is unavailable or unnecessary. By contrast, in RYGB with long limbs, standard colonoscopes and therapeutic gastroscopes are usually inadequate due to limited working length. In such settings, BAE is generally required, with reported technical success rates of 70%-90% depending on operator experience and limb length[12]. In post-Whipple surgery, the main challenge lies in identifying the pancreatobiliary limb among adhesions and multiple anastomoses, whereas in hepaticojejunostomy strictures, deep small-bowel intubation is often necessary, making balloon-assisted techniques the preferred option[13].

In this context, device-assisted enteroscopy (DAE) plays a key role in performing ERCP in patients with SAA. In recent years, DAE has largely replaced push enteroscopy, enabling diagnostic and therapeutic procedures to be performed along the entire length of the SB. DAE is a generic term that describes any endoscopic technique for SB examination that involves assisted progression, and it includes procedures such as double-balloon enteroscopy (DBE), single-balloon enteroscopy (SBE), and ultimately motorized spiral enteroscopy (MSE)[11]. Consequently, DAE represents the gold standard endoscopic procedure for SB explorationl[14] and enteroscopy-assisted ERCP has demonstrated high success rates[15,16]. However, there is the limited availability of specific devices for DAE-ERCP, the need for skilled personnel, and the high cost and time of the procedure.

Balloon-assisted enteroscopy-ERCP: Single and double balloon

Biliary endoscopy in patients with altered anatomy requires not only a comprehensive understanding of the various surgical procedures and upper GI tract modifications, but also high technical expertise and thorough knowledge of available endoscopes and devices[17].

The overall success rate of ERCP in patients with altered anatomy ranges from 70% to 90%[18], depending on the type of surgical reconstruction and the endoscopist expertise. BAE, which includes SBE and DBE[18], allows endoscopic insertion into the deeper portion of the SB, thus providing access to the biliary system in patients with SAA. ERCP procedures using SBE or DBE have positive results in the literature despite difficulties related to advancing the instrument till the site of interest and performing the cholangiopancreatography itself. These procedures require an overtube with a balloon attached to its tip, which is placed over the endoscope (single balloon), while a second balloon at the distal end is provided for DBE procedures. A "push-and-pull" technique is combined with an inflation-deflation method used in DBE procedures to gradually pleat and shorten the SB over the overtube to allow endoscope progression[19]. The SBE procedure combines the "push-and-pull" technique with a "hook-and-suck" method in which the endoscope is held in a stable position by tip angulation and suction on the bowel wall[14].

Successful advancement through complex afferent limbs in DAE-ERCP relies on coordinated maneuvers such as loop minimization, stable overtube management, and fluoroscopic guidance. Reducing loops - by carefully withdrawing the scope with torque, assisted by abdominal pressure or the push-and-pull technique - facilitates deeper insertion. Secure overtube positioning is equally important to prevent loss of stability and mucosal trauma. Fluoroscopy provides complementary orientation, confirming limb direction and helping to recognize and correct looping. Familiarity with these strategies significantly improves technical success in altered anatomy ERCP[18,20].

Recently, shorter DBE and SBE enteroscopes have been introduced, providing improved maneuverability and enabling the use of conventional ERCP accessories and catheters (Figure 2). Thanks to their short length and 3.2 mm working channel, both types of enteroscopes allow the use of the standard devices employed during ERCP, including sphincterotomes, stone extraction accessories, and both plastic and self-expandable metal stents (SEMS)[13,21].

Figure 2  Fluoroscopic image of short-single-balloon enteroscopy assisted endoscopic retrograde cholangiopancreatography in a Roux and Y reconstruction.

Several studies, including retrospective analyses and meta-analyses, indicate that both SBE and DBE achieve similar overall success rates for ERCP in SAA. Intubation, cannulation, and therapeutic success rates are mostly comparable between the two techniques. A recent multi-center study of short SBE-assisted ERCP in patients with SAA reported a total procedural success and AE rate of 74.9% and 7.7%, respectively[22]. The same study identified several factors associated with procedural failure, including Roux-en-Y anastomosis and malignant cases. On the other hand, a multicenter real-world study by Farina et al[19] observed an overall technical success rate of 86% for DBEERCP in 67 procedures among 53 patients with SAA, noting progressive improvement in outcomes over time and underscoring the importance of experience and referral-center volume.

While both methods are technically demanding and time-consuming, some evidence suggests that average procedural times for SBE-ERCP might be slightly shorter than for DBE-ERCP[16]. A point to consider is that long SBE enteroscope features a working length of 200 cm requiring longer accessories, a problem that the use of a "short" double- or single-balloon enteroscope could solve in some patients, however, the adoption of short SBE scopes remains limited, as these instruments are still relatively recent innovations. A recent study by Cho et al[23] proposed a scope-exchange technique to overcome the limitations of the long SBE enteroscope. After advancing the enteroscope and leaving its overtube in place, a conventional gastroscope was inserted through the overtube to perform ERCP with standard instruments. The method was effective mainly in native papilla anatomies (success > 80%), but failed in bilioenteric anastomoses (0% success) due to angulations and overtube instability. Main risks included loss of position, overtube kinking, and mucosal injury, with one major perforation (1.8%) reported. These findings stress the need for careful patient selection.

In conclusion, both double- and single-balloon DAE-ERCP are comparable in diagnostic and therapeutic efficacy in patients with altered anatomy, but SBE may offer an advantage in terms of procedural time[24-26].

BAE-ERCP: AEs

Complications of BAE-ERCP in patients with surgically altered GI anatomy represent a critical aspect of the procedure, although published data on their specific incidence remain variable and, at times, conflicting due to differences in study design, patient populations and reporting standards. The difficulty in reaching and manipulating biliopancreatic structures through an altered anatomy can potentially increase the risk of AEs, making complication management more complex and requiring careful pre-procedural patient evaluation. The overall AE rate ranges from 5% to 10%, with pancreatitis, perforation, and bleeding being the most reported complications[27]. A more recent systematic review and metaanalysis by Gkolfakis et al[28] involving over 600 enteroscopy-assisted ERCP found a pooled complication rate of approximately 5.7%, with similar rates among SBE and DBE techniques. Most reported complications included pancreatitis, perforation, cholangitis, and bleeding, with perforations being rare but potentially serious, especially in Billroth II or Roux-en-Y anatomies. In general, most AEs following BE-ERCP are managed conservatively, though surgical intervention may be required in isolated cases[28]. A specific complication associated with the mechanical aspects of the procedure in complex anatomies, is the development of severe subcapsular hepatic hematoma[29]. This phenomenon has been attributed to the combination of a "closed loop" mechanism and an excessively deep introduction of the guidewire into the intrahepatic biliary tree[30]. This high pressure, combined with possible trauma to the peripheral liver parenchyma by the guidewire (even without capsular rupture), can cause capsular dehiscence and bleeding[27]. To prevent such barotrauma, intermittent desufflation of the enteroscope's and overtube's balloons is recommended to release intraluminal pressure, in addition to the use of CO₂ insufflation.

MSE-assisted ERCP: A short experience

MSE is an advanced endoscopic technique that uses a motorized overtube with a spiral fin system (PowerSpiral, Olympus^®) to pleat the SB onto the enteroscope shaft, enabling rapid and deep insertion with continuous rotation controlled by a foot pedal. Unlike balloon-assisted systems, MSE provides both diagnostic and therapeutic access with a shorter learning curve and faster advancement, especially useful in altered anatomy[31,32].

MSEassisted ERCP has been studied for biliopancreatic interventions in post-surgical patients. A single-center retrospective cohort between 2016 and 2021 involving 36 patients (30 Roux-en-Y, 6 Billroth II) reported technical success in 86.1%, cannulation success in 83.9%, and therapeutic intervention in all cannulated cases, yielding an overall ERCP success rate of 72.2% with one major complication (2.8%)[31]. A subsequent prospective multicenter registry involving multiple European tertiary centers from 2022 to mid-2023 revealed that MSE-assisted ERCP achieved only 54% overall technical success, prompting premature study termination due to safety concerns, including one esophageal perforation tied to device use[33]. Despite its potential, motorized SE was officially withdrawn from the global market in July 2023 following an urgent safety notice after a fatality caused by device impaction and failure to withdraw, with additional reports of required surgical removal due to severe AEs[32]. Despite promising initial reports, enthusiasm for MSE waned for reasons beyond the urgent safety notice. While early single-center studies suggested high technical and therapeutic success, larger multicenter experiences showed lower reproducibility and higher complication rates. Additional limitations included difficulties in scope control within SAA, a steeper-than-expected learning curve, and variability in operator outcomes across centers. The restricted availability of the device, high acquisition costs, and the continued reliability of BAE - with a longer track record of safety and efficacy - further limited widespread adoption[33,34]. Ultimately, MSE was officially withdrawn from the global market in July 2023 after a fatal device-related complication, consolidating its decline in clinical practice[32].

EUS-BD IN SAA PATIENTS

As previously mentioned, surgical anatomy alterations such as RYGB and hepaticojejunostomy often present considerable difficulties for ERCP due to the challenge of accessing the biliary system[35,36]. Historically PTBD or surgical interventions were the alternatives for endoscopic BD in complex cases[37]. However, with the advancements in endoscopic techniques and technologies, EUS-BD has emerged as a viable and less invasive option for managing biliary obstruction in patients with SAA, where enteroscopy-assisted ERCP fails or is not feasible, with high technical and clinical success rates[38,39].

An Italian retrospective study recently demonstrated wide variability in the endoscopic management of BD in patients with SAA. Notably, it reported a rising preference for EUS-BD, especially in cases involving Roux-en-Y anatomy, where it appears to offer higher clinical success rates than DAE. Despite this trend, the survey revealed that complex cases are infrequently referred to specialized centers, and procedural volumes remain low in non-tertiary institutions. These findings highlight the need for more consistent clinical pathways, centralized care for complex cases, and stronger inter-institutional collaboration[40].

EUS-BD involves accessing the biliary system under direct ultrasound guidance, typically from the stomach or duodenum, depending on the location of the obstruction and the altered anatomy[41]. This approach allows for targeted intervention, avoiding the need for percutaneous access[42].

EUS-BD can be performed by four approaches (Table 1): EUS-rendezvous (EUS-RV), EUS-transluminal BD, EUS-guided antegrade intervention (EUS-AI) and EUS directed transgastric ERCP (EDGE) procedure[43].

Table 1 Different types of surgical altered anatomy and the corresponding endoscopic ultrasound-guided biliary drainage procedures.

Surgically altered anatomy	EUS biliary drainage procedure
Sleeve gastrectomy	EUS-CD, rendezvous, EUS HGS, antegrade drainage
Billroth-I gastrectomy	EUS-CD, rendezvous, EUS-HGS, antegrade drainage
Billroth-II gastrectomy	EUS-HGS, rendezvous, antegrade drainage
Roux-en-Y gastric bypass	EUS-HGS, antegrade drainage, EDGE
Whipple’s procedure	EUS-HGS, rendezvous, antegrade drainage
Hepaticojejunostomy	EUS HGS, rendezvous, antegrade drainage

EUS: Endoscopic ultrasound; EDGE: Endoscopic ultrasound directed transgastric endoscopic retrograde cholangiopancreatography; EUS-HGS: Endoscopic ultrasound-guided hepatogastrostomy; EUS-CD: Endoscopic ultrasound-guided choledochoduodenostomy.

EUS-RV: In patients with SAA - such as RYGB, Billroth II gastrectomy, or hepaticojejunostomy - EUS-RV serves as a valuable rescue technique. The method involves EUS-guided puncture of an intrahepatic bile duct (usually segment B2 or B3), passage of a guidewire through the biliary-enteric anastomosis or ampulla, and subsequent wire retrieval with an enteroscope or duodenoscope to complete the ERCP. EUS-RV has been successfully employed in patients after pancreaticoduodenectomy, including Whipple procedures, allowing access to the biliary system despite the complex postoperative anatomy[44]. EUS-RV is the preferred technique in benign biliary obstruction[43].

EUS-guided transluminal BD (EUS-TL): The EUS-TL approach involves creating a direct communication between the biliary system and the GI lumen, typically the stomach or duodenum. After accessing the bile duct under EUS guidance, a fistula is created between the bile duct and the GI lumen through a lumen-apposing metal stent (LAMS). There are two transluminal techniques: Hepatogastrostomy and choledochoduodenostomy[45]. EUS-guided choledochoduodenostomy (EUS-CD) involves creating an anastomosis between the common bile duct and the duodenum under EUS guidance. This technique is typically performed in patients with distal biliary obstruction where access to the papilla is not feasible[46].

EUS-guided hepaticogastrostomy involves creating an anastomosis between the left hepatic duct and the stomach under EUS guidance. This approach is often used for high biliary obstructions or when access to the distal bile duct is not possible, as in the anatomical alterations reported above. Following transmural access to the bile duct, the tract is progressively dilated to create a mature fistulous tract between the biliary tree and the GI lumen, allowing the deployment of a fully or partially covered self-expandable metal stent to achieve stable extra-anatomical BD[47]. The intrahepatic approach is often favored in SAA due to anatomical orientation and access stability. Studies report high technical and clinical success rates - often exceeding 85%-90%. The advantages in SAA of this approach is particularly useful when the ampulla is inaccessible, e.g., Roux-en-Y or Whipple, where duodenoscope access is nearly impossible Furthermore Transgastric approach is feasible regardless of distal bowel anatomy. While the limitations are: (1) Technically challenging in non-dilated intrahepatic ducts; (2) Higher risk of complications like bile leak, peritonitis, or stent migration; and (3) Requires left-sided biliary obstruction for best outcomes. Meta-analyses and large series report overall AE rates ranging from 15%to 33% with severe or fatal events occurring in a minority of cases (severe 3%-4%, fatal 3%)[48-50]. Stent related complication such us recurrent biliary obstruction (RBO) and stent migration are also notable, with RBO rates up to 33% in long term follow-up. But by choosing the ideal stent according to the needs of the maneuver, the risks related to migration or obstruction of the stent can also be reduced. In this regard LAMS is ideal for short-term use and rapid fistula formation, especially in EUS-CDS, but should be removed within 2-3 months to avoid complications. Partially covered-SEMS is better suited for longer-term BD, especially in EUS-HGS, with moderate migration risk that can be mitigated by proper placement and design. In summary most AEs are mild to moderate and can be managed conservatively or with endoscopic reintervention[48,49].

EUS-AI: EUS-guided antegrade approach is a technique that involves placing a stent through the obstructed biliary segment under EUS guidance. After accessing the bile duct, a guidewire is advanced through the obstruction, and a stent is deployed to maintain BD. This approach is particularly useful for managing hilar biliary obstructions or when conventional ERCP is not feasible due to the location or nature of the obstruction[51].

EDGE: EUS-directed transgastric ERCP is a minimally invasive endoscopic technique developed to facilitate ERCP in patients with SAA, most commonly those with RYGB. In RYGB, the native duodenum and ampulla are inaccessible via standard endoscopy, making conventional ERCP technically challenging or sometimes not feasible. Under EUS guidance, a LAMS is placed to create a temporary access from the gastric pouch to the excluded stomach (step 1). This allows a standard duodenoscope to pass through the fistula into the excluded stomach and reach the ampulla (step 2)[35,43]. These two steps may be executed either within a single endoscopic session or in a staged approach (Figure 3). Performing both stages in a single session enables immediate biliary intervention, but increases the risk of LAMS dislodgement and consequent perforation, due to insufficient tract maturation. Tract maturation refers to the process by which a stable, fibrous channel forms between two luminal structures (e.g., gastric pouch to excluded stomach or jejunum) following LAMS placement. A mature tract significantly reduces the risk of leakage, peritonitis, or stent migration during subsequent endoscopic interventions (e.g., ERCP). Maturation allows for safer passage of larger-caliber endoscopes without disrupting the fistula. Therefore a dual-session approach, with a 10-14 days interval, allows for fistula stabilization, significantly reducing the risk of AEs, albeit at the cost of delayed therapy[52]. A retrospective study compared single-session vs shortened-interval dual-session EDGE in patients with RYGB, showing that a 2-4 days interval between LAMS placement and ERCP significantly reduced the risk of stent dislodgement without delaying therapy. The findings support the dual-session approach as a safer and effective strategy[53]. The American Society for Gastrointestinal Endoscopy specifically recognizes EDGE as an appropriate and effective approach for biliary access in patients with RYGB, with technical and clinical success rates comparable to laparoscopic-assisted ERCP (LA-ERCP) and superior to enteroscopy-assisted ERCP in terms of procedural efficiency and access[54]. Furthermore, in order to minimize possible AEs, the correct anchoring of the LAMS is essential and is possible through research of a secure positioning of the LAMS flanges against the luminal walls to prevent dislodgement or migration. Using a LAMS of appropriate diameter - typically 15-20 mm - is recommended to facilitate easy passage of the endoscope while ensuring secure tissue apposition. Proper anatomical alignment between the gastric pouch (such as in RYGB patients) and the remnant stomach or jejunum is also crucial during LAMS placement. Pre-procedure imaging - such as computed tomography scans or fluoroscopy - combined with real-time EUS assessment of remnant stomach distensibility, is essential for choosing the optimal puncture site. Patient positioning (e.g., left lateral or supine) may further optimize the alignment between the pouch and remnant stomach, enhancing procedural safety and success[55].

Figure 3  Endoscopic ultrasound directed transgastric endoscopic retrograde cholangiopancreatography.

EUS-BD has demonstrated high technical and clinical success rates in patients with SAA. However, EUS-BD is associated with potential complications including bleeding, bile leakage, peritonitis, cholangitis, perforation, and stent migration or obstruction. The reported rates of overall AEs range from 6% to 20%. When compared directly with PTBD, EUS-BD shows higher clinical success rates, fewer AEs and reinterventions, along with shorter hospital stays and lower costs[41,56]. EUS-BD demonstrates higher technical and clinical success than device-assisted ERCP in patients with SAA, and is associated with shorter procedure times. A recent review by Vanella et al[42] concluded that EUS-BD, particularly in expert hands, is transitioning from a rescue technique to a standard minimally invasive option, though wider adoption will require dedicated devices, structured training, and prospective multicenter validation. However, it carries a higher rate of AEs and typically requires advanced expertise and center experience. According to the European Society of Gastrointestinal Endoscopy (ESGE) guidelines, EUS-BD is recommended as a valid alternative to DAE-ERCP in cases of altered anatomy, especially after failed ERCP or when DAE expertise is not available, emphasizing the need for individualized treatment planning based on patient anatomy, local expertise, and procedural risk[45].

Last but not least, EUS-BD requires structured training and credentialing to ensure safety and efficacy. Current ASGE recommendations suggest that only endoscopists proficient in both diagnostic EUS (≥ 50 procedures with FNA) and therapeutic ERCP (≥ 180 cases) should attempt EUS-BD, given its technical complexity[57]. Proficiency may require at least 33 EUS-hepaticogastrostomy cases to achieve the learning curve plateau[58]. Training should progress from diagnostic EUS to simpler interventional procedures before advancing to EUS-BD, ideally within accredited centers equipped with multidisciplinary support[35]. Even today though credentialing for EUS-BD is not universally standardized, but should include formal documentation of advanced endoscopy fellowship or equivalent experience, with institutional privileging based on demonstrated procedural volume and outcome.

ENDOSCOPY DRAINAGE IN SAA: ANATOMY-SPECIFIC APPROACHES

Billroth II gastrectomy

In Billroth II anatomy, the afferent limb is usually short, and access to the papilla is feasible without balloon enteroscopy. Operative gastroscopes or pediatric colonoscopes are particularly useful in this setting: Comparative studies show that ERCP success rates with pediatric colonoscopes in Billroth II are high, with shorter procedure times compared to balloon-assisted approaches. A retrospective study[59] identified previous ERCP, absence of Braun anastomosis, and use of a cap-assisted gastroscope as predictors of success, confirming the safety and efficacy of forward-viewing endoscopes in this setting. In a more recent retrospective study the technical success in BII patients was 97%, consistently showing the highest technical success rates compared with RYGB and other SAA patients[12]. When standard ERCP fails, EUS-RV approach offers a reliable rescue strategy. In large series, EUS-BD techniques (including rendezvous, antegrade and hepatico-gastrostomy) achieved technical success of 89.4% and clinical success of 96.2%, with almost 97.4% clinical success specifically in the subset of patients with proximal stenoses or SAA - clearly superior to percutaneous drainage in terms of success and patient recovery[60].

RYGB

As already mentioned, among the different types of postsurgical anatomy encountered in clinical settings, Roux-en-Y gastrectomy is one of the most difficult reconstructions. ERCP in these patients is a greater challenge not only due to the long afferent limb of the Y anastomosis often difficult to identify, but also due to the presence of the native papilla, which could be difficult to cannulate[29]. In Roux-en-Y anatomy, recognizing the afferent limb can be improved by using intraluminal indigo carmine dye during balloon-assisted ERCP - this technique helps distinguish the correct limb by visualizing dye uptake[61], while the “tidal wave sign”, characterized by retrograde peristalsis in the afferent limb, accurately identifies it in about 84% of cases[62]. The papilla of Roux-en-Y gastrectomy is difficult to cannulate due to the reverse position of the papilla, difficulty with scope manipulation and improper accessories (lack of an elevator)[17]. A recent large retrospective study involving 334 patients (665 ERCPs) with altered anatomy reported enteroscopy success rates of 82.2% in RYGB, with intervention success of 65.1% in RYGB - significantly lower than other anatomies[12]. Complications are not more frequent per se, but procedure length and looping predispose to increased patient discomfort and rare events such as hepatic subcapsular hematoma from guidewire trauma[27]. In this context, EUS-guided approaches have revolutionized management. The EDGE procedure is increasingly recognized as the preferred method. Alternatively, when EDGE is not feasible, EUS-guided transluminal drainage techniques, such as choledochoduodenostomy or hepaticogastrostomy, may be used. These approaches have demonstrated near-100% technical success and clinical success above 80%, positioning EUS-BD as a superior alternative to device-assisted ERCP in this group[40,60].

Post-Whipple reconstruction and hepaticojejunostomy

In post-Whipple anatomy, access to the bilioenteric anastomosis is usually possible with either pediatric colonoscopes or short enteroscopes, depending on limb length. Short DBE and SBE enteroscopes are advantageous as their 3.2 mm channel accommodates standard ERCP devices, reducing the need for specialized accessories[21,22]. In a recent retrospective series including 106 SBE-ERCP procedures in 46 post-Whipple patients, technical success for biliary indications reached 90%, with clinical success of 88%. In contrast, pancreatic indications were more challenging, with clinical success limited to 65%. AEs, mainly related to cannulation difficulty and anastomotic strictures, occurred in 11% of biliary cases, all mild, and no severe complications were observed[63]. Patients with hepaticojejunostomy, however, present a particularly challenging scenario due to the long afferent loop and frequent need for repeated therapeutic interventions. In this setting, DAE-ERCP has reported success rates ranging from 70% to 85%, but AEs are more frequent compared to Whipple or Billroth II, with bleeding and perforation being the most relevant[28]. Here, scope selection is critical, with short DBE/SBE enteroscopes offering the best balance between reach and accessory compatibility. In patients with other forms of SAA where the papilla or anastomosis is accessible, intrahepatic EUS-RV is performed (and extrahepatic EUS-RV may be considered if the extrahepatic route is available). If EUS-RV is not feasible or successful, or if the papilla/anastomosis is not accessible, EUS-HGS and antegrade intervention are considered[35].

A schematic overview of the stepwise approach to SAA management is provided in the flowchart (Figure 4).

Figure 4 Flowchart for biliary drainage in surgically altered anatomy. ERCP: Endoscopic retrograde cholangiopancreatography; EA-ERCP: Enteroscopy-assisted endoscopic retrograde cholangiopancreatography; EUS-BD: Endoscopic ultrasound-guided biliary drainage; EDGE: Endoscopic ultrasound directed transgastric endoscopic retrograde cholangiopancreatography; HGS: Hepatogastrostomy; PTBD: Percutaneous transhepatic biliary drainage; LA-ERCP: Laparoscopy-assisted endoscopic retrograde cholangiopancreatography.

NON-ENDOSCOPIC ALTERNATIVE TECHNIQUES

PTBD in SAA

When both DAE-ERCP and EUS-BD are unsuccessful or unavailable in SAA patients, PTBD serves as an effective salvage intervention[63]. The PTBD procedure involves inserting a catheter through the liver into the biliary system for external drainage or subsequent stent placement. Pre-procedural imaging is essential to delineate anatomy and select the optimal access route based on surgical reconstruction and obstruction level. After bile duct puncture, guidewire insertion, and tract dilation, an external drainage tube is initially placed. Typically, within about a week, a covered self-expandable metal stent (CSEMS or UCSEMS) is deployed at the obstruction to replace the plastic tube, allowing for durable internal drainage[64]. Studies have reported higher technical success rates with PTBD (approximately 90%) compared to DAE-ERCP (approximately 56%-70%), especially in complex anatomies such as Roux-en-Y or post-liver transplant biliary-enteric anastomoses. However, PTBD is associated with more frequent long-term interventions, pain, catheter-related complications, and inferior quality-of-life due to external drainage systems. Compared to EUS-BD, PTBD also requires more frequent reinterventions and results in higher late AE rates (e.g., 53.8% vs 6.6%), despite similar technical success (approximately 96%-97%)[42,63,64]. From a cost-effectiveness standpoint, EUS-BD may involve higher upfront costs due to the need for specialized equipment and expertise, but multiple studies demonstrate that it is more cost-efficient in the long term, with fewer complications, reduced reinterventions, and shorter hospital stays. Cost analyses from both US and international systems consistently show moderate to substantial savings with EUS-BD compared to PTBD, largely driven by its improved safety profile[65,66]. The American Society for Gastrointestinal Endoscopy highlights EUS-BD as the preferred rescue approach after failed ERCP in centers with adequate expertise, noting its “moderate savings” over PTBD[54]. The American College of Gastroenterology also notes that EUS-BD may be preferred over PTBD due to similar efficacy but fewer complications and reinterventions, though emphasizes that these procedures should be performed by highly experienced endoscopists[67]. Similarly, the American College of Gastroenterology recommends EUS-BD over PTBD for similar efficacy with fewer AEs, stressing it should only be performed by highly experienced endoscopists. PTBD remains a fallback when EUS expertise is unavailable but is associated with higher complication, reintervention, and resource use rates[66].

Table 2 shows the comparative outcomes between entero-ERCP, EUS, and PTBD and Table 3 indicates the BD strategies in patients with SAA.

Table 2 Comparative outcomes table between entero-endoscopic retrograde cholangiopancreatography, endoscopic ultrasound, and percutaneous transhepatic biliary drainage.

Modality	Billroth II gastrectomy	RYGB	Post-Whipple and hepaticojejunostomy
Enteroscopy-assisted ERCP	Technical success: High, up to 97%. Clinical success: High. AE rates: Low. Procedural time: Shorter than other SAA, often using forward-viewing scopes	Technical success: 82.2%. Clinical success: Lower, 65.1% (intervention). AE rates: 5.7% (overall). Rare complications like hepatic subcapsular hematoma. Procedural time: Long and technically demanding due to long, tortuous limb	Technical success: 70%-90%. Clinical success: 88% (biliary). AE rates: Up to 11% (biliary), generally mild. Procedural time: Long and complex
EUS-guided biliary drainage	Technical success: Very high, 89.4% (for all EUS-BD methods). Clinical success: Very high, 96.2%. AE rates: 6%-20%. Procedural time: More efficient than failed ERCP attempts	Technical success: > 85%-90%, with EDGE reaching > 97.9%. Clinical success: > 80%. AE rates: 13.1% (for EDGE). Procedural time: Shorter and more efficient than enteroscopy, particularly for long-limb reconstructions	Technical success: 80%-90%. Clinical success: 85%-92%. AE rates: 10%-20%. Procedural time: More efficient as a rescue method
PTBD	Technical success: 90%-97% clinical success: High AE rates: < 15% (long term issues and reinterventions) procedural time: 1-2 hours

Technical success: Refers to achieving the technical goal (e.g., stent placement); clinical success: Refers to the resolution of the patient's symptoms (e.g., reduction of cholangitis); adverse event rate: The percentage of reported complications. The overall rates are summaries from the cited literature and can vary significantly depending on the specific study. ERCP: Endoscopic retrograde cholangiopancreatography; AE: Adverse event; SAA: Surgically altered anatomy; EUS: Endoscopic ultrasound; EUS-BD: Endoscopic ultrasound-guided biliary drainage; EDGE: Endoscopic ultrasound directed transgastric endoscopic retrograde cholangiopancreatography; PTBD: Percutaneous transhepatic biliary drainage; RYGB: Roux-en-Y gastric bypass.

Table 3 Biliary drainage strategies in patients with surgically altered anatomy.

Type of anatomy	Preferred approach	Indicated instruments	Alternative approaches
Billroth II gastrectomy	ERCP with forward-viewing scope	Operative gastroscope, pediatric colonoscope	EUS-rendezvous
Roux-en-Y gastric bypass	EUS-BD (EDGE)	EUS endoscope, LAMS stent	PTBD, surgical drainage, LA-ERCP
Post-Whipple and bilioenteric anastomoses	DAE-ERCP	SBE, DBE	EUS-rendezvous, EUS-BD, PTBD

ERCP: Endoscopic retrograde cholangiopancreatography; EUS-BD: Endoscopic ultrasound-guided biliary drainage; EDGE: Endoscopic ultrasound directed transgastric endoscopic retrograde cholangiopancreatography; EUS: Endoscopic ultrasound; PTBD: Percutaneous transhepatic biliary drainage; LAMS: Lumen-apposing metal stent; LA-ERCP: Laparoscopic-assisted endoscopic retrograde cholangiopancreatography; DAE: Device-assisted enteroscopy; SBE: Single-balloon enteroscopy; DBE: Double-balloon enteroscopy.

Surgical drainage in SAA

When endoscopic methods fail for biliary access in altered anatomy - particularly long-limb Roux-en-Y or post-Whipple reconstructions - surgical drainage remains a salvage option. Open or laparoscopic surgical options include bilioenteric bypasses (e.g., hepaticojejunostomy) or stone removal via transenteric choledochotomy. LA-ERCP is a valuable technique for managing biliary issues in patients with altered anatomy, such as those post-RYGB. A systematic review and meta-analysis by Gkolfakis et al[28] found that LA-ERCP achieved the highest technical success rate at 99.1%, surpassing EDGE at 97.9% and DAE-ERCP at 87.3%. However, LA-ERCP was associated with a higher incidence of AEs (15.1%) compared to EDGE (13.1%) and EA-ERCP (5.7%). Notably, combined endoscopic-percutaneous rendezvous approaches and internalization strategies are increasingly important in tertiary practice, allowing staged conversion from external to internal drainage and facilitating access in complex reconstructions[68]. Careful risk-benefit assessment and performance in experienced centers are recommended for this approach[28].

ENDOSCOPIC CHALLENGE AND FURTURE

In patients with SAA, BD remains a complex endoscopic challenge requiring a tailored approach based on anatomical reconstruction, local expertise, and procedural risk. While significant advances have been made, there is potential for further improvement. When determining the most appropriate endoscopic technique for BD in patients with SAA some factors should be considered: First, an alimentary limb longer than 150 cm makes reaching the papilla very difficult with BAE-ERCP. Further, in urgent settings, BAE or PTBD may be safer choices than single-session EDGE which has often been associated with complications[69]. Equally important, procedural success depends not only on the choice of technique but also on mastering its specific tips and tricks: For example, in enteroscopy-assisted ERCP, the use of a transparent cap and performing the procedure underwater have been shown to enhance visualization and stability[70]. Familiarity with such technical refinements, as highlighted in the literature, is essential to optimize outcomes.

CONCLUSION

In conclusion, current evidence supports a multidisciplinary, step-up strategy in expert centers, and highlights the urgent need for standardized algorithms, increased referral to high-volume institutions, and further comparative studies to define the most effective and safest BD modality in this challenging setting.