Efficacy and safety of double-balloon enteroscopy-assisted endoscopic retrograde cholangiopancreatography in surgically altered anatomy: A systematic review and meta-analysis

Abstract

BACKGROUND

Performing endoscopic retrograde cholangiopancreatography (ERCP) in patients with surgically altered gastrointestinal anatomy poses significant challenges. Double-balloon enteroscopy-assisted ERCP (DBE-ERCP) has emerged as a safe and effective approach in this patient population. This study aims to provide an updated systematic review and meta-analysis of the safety and efficacy of DBE-ERCP in patients with surgically altered anatomy (SAA), building on previously published evidence.

AIM

To evaluate the safety and efficacy of DBE-ERCP in patients with SAA through an updated systematic review and meta-analysis and to compare outcomes between short- and long-scope double-balloon enteroscopy (DBE).

METHODS

A comprehensive search of PubMed, EMBASE, and Web of Science was performed for studies published up to March 2025 investigating DBE-ERCP in patients with surgically altered gastrointestinal anatomy. A random-effects model was applied to conduct a meta-analysis of proportions. The risk of bias was evaluated using the Newcastle-Ottawa Scale and the Joanna Briggs Institute Scale. Heterogeneity was evaluated using the inconsistency statistic (I²). Publication bias was examined using funnel plots and Egger’s regression test.

RESULTS

A total of 40 studies were included, comprising 10 cohort studies and 30 case series, including 2689 patients who underwent 3478 procedures. The surgical procedures were primarily classified into three categories: Roux-en-Y reconstruction (including hepaticojejunostomy, gastric bypass, and choledochojejunostomy, etc.) in 1156 cases; pancreaticoduodenectomy (performed using either the Whipple or Child technique) in 549 cases; and Billroth II anastomosis in 265 cases. The combined success rate for reaching the papilla was 92% (95%CI: 89%-95%). The overall enteroscopy success rate was 89% (95%CI: 85%-92%). The pooled diagnostic success rate was 90% (95%CI: 85%-95%), while the therapeutic success rate reached 92% (95%CI: 89%-95%). Adverse events reported in 5.7% of patients (95%CI: 4.1%-7.5%). Subgroup analysis comparing short-scope and long-scope demonstrated that the short DBE was superior in terms of papilla reached rate, enteroscopy success, and procedural success. No significant differences were observed between groups in diagnostic success or adverse events.

CONCLUSION

DBE-ERCP demonstrates both safety and efficacy in patients with SAA. Compared to long-scope DBE, short-scope DBE shows greater clinical promise; however, further randomized controlled trials are warranted to validate these findings.

Key Words: Double-balloon enteroscopy; Single-balloon enteroscopy; Surgically altered anatomy; Double-balloon enteroscopy-assisted endoscopic retrograde cholangiopancreatography; Roux-en-Y; Billroth II

Core Tip: Double-balloon enteroscopy-assisted endoscopic retrograde cholangiopancreatography (DBE-ERCP) is a safe and effective approach for the management of pancreatobiliary diseases in patients with surgically altered anatomy. This updated systematic review and meta-analysis of 40 studies including 2689 patients demonstrates high enteroscopic (89%), diagnostic (90%), and procedural (92%) success rates, with a low overall adverse event rate (5.7%). Subgroup analysis indicates that short-type double-balloon enteroscopy (DBE) outperforms long-type DBE in procedural success and papilla access. These findings highlight the clinical utility of DBE-ERCP and suggest that short-type DBE may offer greater advantages, though further randomized controlled trials are warranted.

INTRODUCTION

Endoscopic retrograde cholangiopancreatography (ERCP) is a crucial technique for the diagnosis and treatment of biliary and pancreatic diseases. In patients with normal anatomy, the procedural success rate can reach 90%-95%[1,2]. However, those with surgically altered anatomy (SAA) and suspected pancreaticobiliary diseases pose unique challenges[3,4]. The technical difficulty of ERCP in SAA patients is mainly due to the elongated digestive tract and sharply angled surgical limbs, which limit endoscope advancement and maneuverability. SAA patients can generally be classified into two types: Gastrojejunostomy and Roux-en-Y enteroenteric anastomosis of the small bowel with an afferent biliopancreatic limb and an efferent alimentary limb. In the first type, it involves Billroth II distal gastrectomy and Whipple duodenopancreatectomy, while in the second type, most surgical reconstructions utilize the Roux-en-Y enteroenteric anastomosis, connecting the biliary and/or pancreatic ducts with the small bowel. Conventional side-viewing duodenoscopes, as well as subsequent attempts with push enteroscopes and pediatric colonoscopes, have demonstrated relatively low success rates[5,6]. In the past, more invasive treatments such as percutaneous interventions or open surgery were considered, but their use was often limited by factors such as anticoagulant therapy, intrahepatic biliary dilatation, or the presence of ascites[7]. Alternative techniques, such as push enteroscopy, device-assisted ERCP [including double-balloon enteroscopy (DBE), single-balloon enteroscopy (SBE), balloon-guided enteroscopy, and spiral enteroscopy], and laparoscopy-assisted ERCP, have significantly enhanced endoscopists' ability to intubate the afferent limb in patients with SAA[8,9]. The device-assisted ERCP facilitate deeper advancement of the endoscope by enabling “pleating” or “telescoping” of the small bowel[10]. Laparoscopy-assisted ERCP is associated with longer procedural times in addition to laparoscopy-related adverse events, including postoperative infections[11]. Percutaneous transhepatic access may also facilitate biliary intervention but often requires multiple procedures and prolonged resolution periods[12]. Compared with laparoscopy-assisted ERCP, EUS-directed transgastric ERCP has a shorter overall procedural time, though the incidence of adverse events is similar[13]. Further studies are still needed to validate these findings. European Society of Gastrointestinal Endoscopy (ESGE) recommends device-assisted enteroscopy (DAE) and ERCP as a first-line endoscopic approach to treat pancreaticobiliary diseases in patients with SAA (except for Billroth II patients)[14]. Since 2005, when Haruta first reported the successful use of DBE-ERCP to treat a Roux-en-Y anastomotic stricture in a 7-year-old boy[15], multiple prospective and retrospective studies have demonstrated its safety and efficacy. However, due to the limited sample sizes in many existing studies, a comprehensive synthesis of data on the use of DBE-ERCP in patients with SAA is still needed. Therefore, building upon previous research, this study conducts a meta-analysis to integrate current evidence and provide an updated systematic evaluation of the efficacy and safety of DBE-ERCP in patients with reconstructed gastrointestinal anatomy.

MATERIALS AND METHODS

This systematic review and meta-analysis has been registered in PROSPERO (registration number CRD420251031845) and was conducted in strict accordance with the PRISMA guidelines.

Search strategy

A comprehensive literature search was conducted using PubMed, EMBASE, and Web of Science for the period from January 2001 to March 2025. We performed a comprehensive search using terms such as “endoscopic retrograde cholangiopancreatography”, “ERCP”, “double balloon enteroscopy”, and “surgically altered anatomy”. The full search strategy is outlined in Supplementary material. To ensure completeness, we also reviewed the bibliographies of included studies and sought input from field specialists to identify any additional relevant articles.

Eligibility criteria

Inclusion criteria: (1) Patients who had undergone gastrointestinal bypass surgery; (2) Patients treated with DBE-ERCP, with analysis of the safety and efficacy of the procedure; (3) Studies that included at least 10 patients; and (4) Both single-arm and double-arm studies were included.

Exclusion criteria: (1) Studies that did not report diagnostic or procedural outcomes; (2) Animal studies or other non-human research; and (3) Review articles, letters, conference abstracts, or editorials.

Data extraction and quality assessment

Two reviewers independently performed data extraction and quality assessment, and conducted data abstraction onto a data collection manual designed a priori. The main information extracted from the original literature included: First author, publication year, country, sample size, baseline characteristics of the patients, interventions, study design type, detailed ERCP procedure information (including the type of double-balloon ERCP performed, procedure duration, indication for the procedure, and therapeutic interventions implemented), and main results (e.g., anastomosis/papilla reached, enteroscopic success, diagnostic success, procedural success, DBE-ERCP-related complications).

Ten cohort studies were assessed using the Newcastle-Ottawa Scale, which categorizes studies across three domains based on eight items, including population selection, comparability, and outcome evaluation[16]. Thirty retrospective studies were assessed using the Joanna Briggs Institute Critical Appraisal Checklist for Case Series, which includes ten items evaluating the quality of case series covering case selection, disease or health problem assessment, and presentation of case data[17]. In case of disagreement between the two reviewers during data extraction or quality assessment, a third reviewer was involved to reach consensus. The quality assessment results are shown in Tables 1 and 2.

Table 1 Quality evaluation of the single-arm studies with JBI scale.

Ref.	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Aabakken et al[34]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Emmett and Mallat[35]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Maaser et al[36]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Mönkemüller et al[37]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Moreels et al[38]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Pohl et al[39]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Shimatani et al[25]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Parlak et al[40]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Cho et al[41]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Mönkemüller et al[42]	Y	Y	Y	Y	Y	Y	Y	Y	N	N
Raithel et al[43]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Chua and Kaffes[44]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Osoegawa et al[45]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Siddiqui et al[46]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Shah et al[8]	Y	Y	Y	Y	Y	Y	Y	Y	N	Y
Varabei et al[47]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Sakakihara et al[48]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Tsutsumi et al[49]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
De Koning and Moreels[6]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Shimatani et al[50]	Y	Y	Y	Y	Y	Y	Y	Y	N	Y
Shimatani et al[7]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Kashani et al[31]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Matsumoto et al[51]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Nishio et al[52]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Sato et al[53]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Sirin and Hulagu[54]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Uchida et al[32]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
CHENG et al[55]	Y	Y	Y	Y	N	Y	Y	Y	N	N
Obata et al[56]	Y	Y	Y	Y	N	Y	Y	Y	N	Y
Sato et al[57]	Y	Y	Y	Y	N	Y	Y	Y	N	Y

(1) Were there clear criteria for inclusion in the case series? (2) Was the condition measured in a standard, reliable way for all participants included in the case series? (3) Were valid methods used for identification of the condition for all participants included in the case series%? (4) Did the case series have consecutive inclusion of participants? (5) Did the case series have complete inclusion of participants? (6) Was there clear reporting of the demographics of the participants in the study? (7) Was there clear reporting of clinical information of the participants? (8) Were the outcomes or follow up results of cases clearly reported? (9) Was there clear reporting of the presenting site(s)/clinic(s) demographic information? (10) Was statistical analysis appropriate? Y: Yes; N: No.

Table 2 Quality evaluation of the eligible studies with Newcastle-Ottawa scale.

Ref.	Selection	Comparability	Outcome
Cheng et al[26]	4	1	3
Tsou et al[58]	4	2	3
Tomoda et al[59]	4	2	2
Liu et al[60]	4	2	2
Tsutsumi et al[61]	4	1	2
Mizukawa et al[62]	4	2	3
Yamada et al[63]	4	2	3
Zamora Nava et al[64]	4	2	2
Farina et al[65]	4	2	3
Yokoyama et al[66]	4	2	2

Definitions

Anastomosis/papilla reached was defined as successful advancement of the enteroscope to the papilla or anastomosis. Enteroscopic success was defined by the ability to cannulate either the papilla or the anastomotic site. Diagnostic and procedural success were defined as successful cholangiography and the completion of required therapeutic interventions, respectively. DBE-ERCP related complications: DBE-ERCP-related complications include cholangitis, pancreatitis, bleeding, and perforation, which need further specific treatment.

Statistical analysis

All data analyses and plotting were performed using R (v.4.3.2 for Windows) and the meta package. The pooled proportions for outcomes such as papilla/anastomosis reached, enteroscopic success, and diagnostic success were calculated using the Freeman-Tukey double arcsine transformation. The Der Simonian and Laird random effects meta-analysis of proportions was used to generate the overall effect size of each outcome. Outcomes are expressed as proportions (percentages) with 95%CI. The pooled analyses are presented as forest plots. We assessed variability across studies by calculating the I² statistic, with values above 50% considered indicative of significant heterogeneity.

RESULTS

Study characteristics

A total of 40 studies met the inclusion criteria and were included for further analysis, of which 10 were cohort studies and 30 were case series. This systematic review and meta-analysis adheres to the PRISMA guidelines[18]. The screening process is illustrated in the flow diagram shown in Figure 1. Table 3 summarizes the detailed characteristics of the included studies and patient demographics. The recruitment period spanned from 2007 to 2024. A total of 3478 procedures in 2689 patients were included in these demographics. Geographically, the majority of studies were conducted in Japan (18/40), followed by Germany (5/40) and the United States (4/40). The surgical procedures performed on the patients included three main categories: Roux-en-Y reconstruction (encompassing hepaticojejunostomy, gastric bypass, and choledochojejunostomy, etc.) in 1156 cases, pancreaticoduodenectomy using either the Whipple or Child technique in 549 cases, and Billroth II anastomosis in 265 cases. Table 4 summarizes the details of the ERCP procedures, including the indications for performing ERCP and the therapeutic measures carried out. Among the included studies, 18 applied DBE-ERCP with a short scope (< 200 cm), 15 used a long scope (200 cm), and 7 incorporated both short- and long-scope approaches. Cholangitis, choledocholithiasis, and anastomotic strictures were the most frequently reported indications for DBE-ERCP. Among the therapeutic measures, stone extraction, stent removal, nasobiliary drainage, and both balloon and biliary duct dilation were most commonly carried out.

Figure 1 Study selection process. SBE: Single-balloon enteroscopy; RAE: Rotational overtube‑assisted enteroscope; SE: Standard enteroscopy.

Table 3 Patient characteristics.

Ref.	Study design	Period	Country	n	Age	Female (%)	Patient anatomy	Inclusion criteria
Aabakken et al[34]	Retrospective case series	2005-2006	Norway	18 procedures in 13 patients	53 (2-81)	31	Roux-en-Y (n = 13)	Patients who underwent DBE-ERCP procedures with RY anatomy
Emmett and Mallat[35]	Retrospective case series	2005-2007	United States	20 procedures in 14 patients	47 (27-73)	50	RYGB (n = 6), PD (n = 4), Liver transplant with HJ (n = 1), Liver transplant with CJ (n = 1), PJ (n = 1), Frey procedure (n = 1)	Patients who had previously had a RYGI, RY Whipple resection with PJ, orthotropic liver transplantation with HJ or CJ, or a Frey procedure who underwent ERCP
Maaser et al[36]	Retrospective case series	2004-2007	Germany	11 procedures in 11 patients	62.8 (30-78)	27	Partial gastrectomy with GI and RY (n = 1), Partial gastrectomy with GI and RY and CD (n = 1), Total gastrectomy with RY (n = 1), Partial gastrectomy with Billroth II GI (n = 2), PD (n = 3), Pylorus-preserving CJ with RY (n = 2), RYHJ (n = 1), Duodenum-preserving pancreatic head resection in combination with CJ (n = 1)	Patients with various anatomic variations in whom DBE-ERCP was performed
Mönkemüller et al[37]	Prospective cohort study	-	Germany	11 procedures in 11 patients	59 (36-77)	9	RYHJ (n = 11)	Patients with RY anastomosis and pancreaticobiliary problems who underwent DBE-ERCP
Moreels et al[38]	Retrospective case series	2006-2008	Belgium	15 procedures	-	-	Roux-en-Y entero-enteric anastomosis	Patients who underwent DBE-ERCP for suspected pathology in the excluded gastrointestinal segment or in the biliopancreatic system after Roux-en-Y entero-enteric anastomosis
Pohl et al[39]	Retrospective case series	2004-2008	Germany	25 procedures in 15 patients	60.2 (25-80)	27	RYCJ (n = 15)	Patients with CJ anastomosis who underwent DBE-ERCP procedures
Shimatani et al[25]	Retrospective case series	2006-2008	Japan	68 procedures in 36 patients	-		RY total gastrectomy (n = 36), Billroth II gastrectomy (n = 17), PD (n = 15)	Patients with pancreaticobiliary disease and previous bowel reconstruction undergoing DBE-ERCP at a single center between January 2006 and December 2008
Parlak et al[40]	Retrospective case series	2006-2008	Turkey	14 procedures in 14 patients	45.3 (28-61)	54	RYHJ (n = 14)	Patients with RYHI who underwent ERCP
Cho et al[41]	Retrospective cohort study	2007-2008	Canada	29 procedures in 20 patients	57.9 (26-85)	50	RYHJ (n = 7), Billroth II (n = 6), RYGJ (n = 5), RYEJ (n = 1), PD with CJ (n = 1)	Patients with small-bowel reconstruction who underwent ERCP using a short DBE
Mönkemüller et al[42]	Retrospective case series	-	Germany	23 procedures in 18 patients	-	-	RYHJ (n = 18)	Patients who underwent DBE-ERCP procedures with Roux-en-Y hepaticojejunostomy
Raithel et al[43]	Prospective case series	2005-2008	Germany	86 procedures in 31 patients	61 (38-84)	54.8	Roux-en-Y reconstruction (n = 30), End-to-side gastrojejunostomy (n = 1)	Patients who underwent DBE-ERCP for pancreatobiliary diseases after complex abdominal anatomy (e.g., Billroth II, Whipple, Roux-en-Y)
Chua and Kaffes[44]	Retrospective case series	2008-2011	Australia	18 procedures in 16 patients	-	-	-	Patients who underwent DBE-ERCP procedures with surgically altered anatomy
Osoegawa et al[45]	Retrospective case series	2006-2011	Japan	47 procedures in 28 patients	74 (54-91)	35.7	Roux-en-Y total gastrectomy (n = 11), Billroth II gastrectomy (n = 15), Roux-en-Y anastomosis with PD (n = 2)	Patients who underwent DBE-ERCP for pancreatobiliary diseases after bowel reconstruction
Siddiqui et al[46]	Retrospective case series	2008-2011	United States	79 procedures in 79 patients	78 (29-86)	62	RYGB (n = 39), Billroth II (n = 3), PD (n = 20), HJ (n = 3), RYHJ (n = 4), RYGJ (n = 5), CJ (n = 2), RYPJ (n = 3)	Patients with various anatomic variations in whom ERCP was performed using short DBE
Shah et al[8]	Retrospective case series	2008-2009	United States	27 procedures in 27 patients	-	-	RYGB (n = 15), Non-RYGB including HJ, Whipple resection, other (n = 12)	Long-limb surgical bypass patients with suspected pancreatobiliary diseases
Varabei et al[47]	Retrospective case series	2009-2012	Belarus	33 procedures in 33 patients	53.7 (38-69)	76	RYHJ, jejunum loop with Braun bypass (n = NR)	Patients with RYHJ anastomoses and jejunum loop with Braun's bypass anastomosis who underwent DBE-ERCP
Cheng et al[26]	Retrospective cohort study	2006-2013	Taiwan	77 procedures in 77 patients	76 (50-95)	24	Billroth II (n = 77)	Patients with B2 gastrectomy in whom standard ERCP failed
Sakakihara et al[48]	Retrospective case series	2008-2012	Japan	44 procedures in 44 patients	63.5 ± 14.0	38.6	Modified Child’s method (n = 27) Roux-en Y reconstruction (n = 17)	Patients who underwent DBE-ERCP procedures with stenosis of choledochojejunal anastomosis
Tsou et al[58]	Retrospective cohort study	2007-2013	Taiwan	47 procedures in 47 patients	54.4 (26-87)	53	RY (n = 47)	Patients with RY anastomosis who underwent DBE-ERC
Tsutsumi et al[49]	Retrospective case series	2007-2013	Japan	72 procedures in 72 patients	69 (37-83)	39	PD (n = 72)	Patients with prior PD including SSPPD and PPPD who underwent attempted diagnostic or therapeutic ERCP
De Koning and Moreels[6]	Retrospective case series	2006-2012	Belgium	30 procedures in 10 patients	60 (30-76)	50	Short-limb RY (< 50 cm) with bilioenteric anastomosis (Whipple or BD) (n = 24), Short-limb RY (< 50 cm) with intact papilla (total gastrectomy) (n = 2), Long-limb RY (> 100 cm) with intact papilla (Scopinaro, GB) (n = 4)	Consecutive patients with RY altered small-bowel anatomy undergoing balloon-assisted ERCP performed by a single endoscopist in a single center
Shimatani et al[50]	Prospective case series	2013-2014	Japan	311 procedures in 311 patients	67.3 (14.6)	32.5	RY (n = 203), Billroth II (n = 26), PD (n = 44), PPPD (n = 31), Others (n = 7)	Consecutive patients above the age of 20 with altered GI anatomies and suspected biliary disease who were referred to eight institutions between June 2013 and May 2014
Tomoda et al[59]	Retrospective cohort study	2001-2014	Japan	20 procedures in 20 patients	50.85 (5-74)	35	RYHJ (n = 18), Not specified (n = 2)	Patients who underwent ERCP using a short DBE upon suspicion of biliary stricture after living donor liver transplantation with HJ reconstruction
Liu et al[60]	Retrospective cohort study	2009-2015	Australia	86 procedures in 52 patients	60.5 (18-89)	25	RYHJ (n = 42), RY with GJ (n = 9)	Patients with RY anastomosis who underwent a DBE-ERCP
Shimatani et al[7]	Retrospective case series	2015-2016	Japan	112 procedures in 100 patients	61.6 (18.2)	64.6	RYHJ (n = 30), RY partial gastrectomy (n = 17), RY total gastrectomy (n = 17), Billroth II gastrectomy (n = 8), PD (n = 13), PPPD (n = 12), Billroth II and RY (n = 1), GJ (n = 1), Jejunal pouch interposition (n = 1)	Procedures performed using new short-type DBE (N-short DBE) in postoperative patients at a single center between August 2015 and April 2016
Tsutsumi et al[61]	Prospective cohort study	2014-2017	Japan	33 procedures in 33 patients	65 (53-73)	56	RYHJ (n = 33)	Patients with suspected biliary diseases who had RYHI with side-to-end jejunojejunostomy
Kashani et al[31]	Retrospective case series	2005-2012	United States	129 procedures in 103 patients	50 (22-82)	87	RYGB (n = 103)	Patients with a history of RYGB who underwent DBE-ERCP
Matsumoto et al[51]	Retrospective case series	2010-2016	Japan	11 procedures in 11 patients	71 (53-74)	37	Child's reconstruction (n = 4) Roux-en-Y (n = 7)	Patients with HJ who underwent DBE-ERC for postoperative bile leakage
Mizukawa et al[62]	Retrospective cohort study	2008-2014	Japan	46 procedures in 46 patients	69 (64-75)	41	PD with Braun anastomosis (n = 29), PD without Braun anastomosis (n = 1), PPPD with Braun anastomosis (n = 12), SSPPD (n = 4)	Patients with suspected benign HJ anastomotic stricture after Whipple's procedure who underwent diagnostic ERCP using a short DBE
Yamada et al[63]	Prospective cohort study	2005-2017	Japan	326 procedures in 326 patients	64.9 (15.5) (EI-580BT cohort)	56	PD without Braun anastomosis (n = 1), PPPD with Braun anastomosis (n = 12), SSPPD (n = 4), Gastrectomy with Billroth II (n = 29), Gastrectomy with RY (n = 82), PD with Billroth II (n = 53), PD with RY (n = 42), RYHJ (n = 96), Liver transplantation with HJ (n = 20)	Patients with surgically altered anatomy that underwent ERCP using double-balloon endoscopy with the EI-580BT or EI-530B endoscope
Nishio et al[52]	Retrospective case series	2011-2018	Japan	104 procedures in 66 patients	70 (66.3-75.8) in the LO group; 67 (61-74) in the control group	44	Liver operation group: Hepatectomy (n = 14), and LDLT (n = 12), Control group: PD (n = 9), PPPD (n = 4), SSPPD (n = 27)	Patients who underwent DBE-ERCP after hepatectomy or LDLT (LO group); Patients who underwent DB-ERCP after pancreatoduodenectomy (control group)
Sato et al[53]	Retrospective case series	2008-2018	Japan	102 procedures in 102 patients	69 (22-84)	40	Roux-en-Y (n = 80), Billroth-II (n = 22)	Patients who underwent DBE-ERCP procedures with Roux-en-Y or Billroth-II for Hepaticojejunostomy anastomotic stricture
Sirin and Hulagu[54]	Retrospective case series	2008-2014	Japan	31 procedures in 31 patients	-	34	Billroth II with long afferenf limb (n = 8), Billroth II with acute angle of anastomosis of the afferent limb (n = 6), Billroth II with R-en-Y (n = 6), RYHJ (n = 11)	Patients with BII gastrojejunostomy and hepaticojejunostomy (with or without Roux-en-Y) who underwent DBE-ERCP during the same session due to failure of standard ERCP using standard duodenoscope
Uchida et al[32]	Retrospective case series	2011-2018	Japan	805 procedures in 319 patients	Median 69 years	38.9	Modified Child method after SSPPD (n = 133), modified Child method after PPPD (n = 29), Roux-en-Y with choledochojejunal anastomosis (n = 116), Roux-en-Y without choledochojejunal anastomosis (n = 19), Billroth II (n = 22)	Patients who underwent DBE-ERCP procedures with surgically altered gastrointestinal anatomy
Zamora Nava et al[64]	Prospective cohort study	2010-2016	Mexico	96 procedures in 75 patients	40.8 ± 10.2	71.9	RYHJ (n = 79), PD (n = 10), Billroth II (n = 3), Gastric bypass (n = 2), Esophagus jejunum anastomosis (Roux-en-Y) (n = 2)	Patients who underwent DBE-ERCP after Predominantly Roux-en-Y hepaticojejunostomy
CHENG et al[55]	Retrospective case series	2015-2020	Hong Kong SAR, China	46 procedures in 37 patients	70.8 ± 15.6	43.2	Billroth II (n = 15), Roux-en-Y (n = 10), RYHJ (n = 5), PD (n = 6), Portoenterostomy (Kasai Operation) (n = 1)	Patients who underwent DB-EERCP with biliary diseases
Obata et al[56]	Retrospective case series	2011-2020	Japan	79 procedures in 79 patients	79 (73-84)	22	Roux-en-Y Gastrectomy (n = 79)	Patients who underwent DB-EERCP procedures with Roux-en-Y for CBD stones
Sato et al[57]	Retrospective case series	2017-2019	Japan	20 procedures in 20 patients	73 (60-75)	45	PD with Billroth-II (n = 10), PD with Roux-en-Y (n = 4), Resection of the extrahepatic bile duct with Roux-en-Y (n = 5), Total pancreatectomy with Billroth-II (n = 1)	Patients who underwent DB-EERCP procedures with HJASs or who were unamenable to balloon dilation or prior plastic stent placement
Farina et al[65]	Retrospective cohort study	2013-2021	Italy	53 procedures in 53 patients	65 (23-89)	40	Pancreaticoduodenectomy (n = 9), Roux-en-Y gastrectomy (n = 17), Roux-en-Y biliodigestive anastomosis (n = 24), Billroth II gastrectomy (n = 3)	Patients who underwent DB-EERCP with post-surgical biliary disorders
Yokoyama et al[66]	Retrospective cohort study	2007-2022	Japan	289 procedures in 289 patients	-	-	RYG (TG) (n = 88), RYG (DG) (n = 33), Cardiac gastrectomy (n = 1), Billroth II (n = 33), RYHJ (n = 38), PD (n = 29), RYLT (n = 35)	Patients who underwent DB-EERCP procedures with SAA treated for biliary stones

ERCP: Endoscopic retrograde cholangiopancreatography; DBE-ERCP: Double-balloon enteroscopy-assisted endoscopic retrograde cholangiopancreatography; RYHJ: Roux-en-Y hepaticojejunostomy; RYGB: Roux-en-Y gastric bypass; RY: Roux-en-Y; HJ: Hepaticojejunostomy; CJ: Choledochojejunostomy; PJ: Pancreaticojejunostomy; GI: Gastrointestinal (or gastrojejunostomy); GJ: Gastrojejunostomy; CD: Choledochoduodenostomy; PD: Pancreaticoduodenectomy; RYCJ: Roux-en-Y choledochojejunostomy; RYEJ: Roux-en-Y esophagojejunostomy; RYGJ: Roux-en-Y gastrojejunostomy; RYPJ: Roux-en-Y pancreaticojejunostomy; RYGI: Roux-en-Y gastrointestinal reconstruction; PPPD: Pylorus-preserving pancreaticoduodenectomy; SSPPD: Subtotal stomach-preserving pancreaticoduodenectomy; LDLT: Living donor liver transplantation.

Table 4 Double-balloon enteroscopy-assisted endoscopic retrograde cholangiopancreatography characteristics.

Ref.	Type of double balloon ERCP	Duration of procedure (minutes)	Indication for ERCP	Therapeutic intervention conducted
Aabakken et al[34]	Long (EN-450 T5, 200 cm)	40 (5-120)		Stent removal (n = 3)
Emmett and Mallat[35]	Long (200 cm)	99 (51-147)	Abnormal LFTs (n = 4), Acute pancreatitis (n = 1), Chronic pancreatitis (n = 2), Cholangitis (n = 2), Recurrent pancreaticobiliary pain (n = 5), Repeat procedure/follow-up (n = 6)	Stone extraction (n = 1), Biliary duct dilation (n = 1), Biliary duct sphincterotomy (n = 2), Biliary duct stent placement (n = 1), Pancreatic duct dilation (n = 1), Pancreatic duct sphincterotomy (n = 1), Pancreatic duct stent placement (n = 1), Pancreatic duct stent removal (n = 1)
Maaser et al[36]	Long (EN-450P5 or EN-450T5, 200 cm)		Choledocholithiasis (n = 1), Cholangitis (n = 4), Biliary pancreatitis (n = 1)	Sphincterotomy (n = 4), Stone extraction (n = 1), Dilation of anastomotic stricture (n = 1)
Mönkemüller et al[37]	Long (EN-450T5, 200 cm)	70 (45-120)	Jaundice (n = 2), Cholestasis (n = 6), Stent extraction (n = 1), CBD-stenosis (n = 1), Choledocolithiasis (n = 1)	Sphincterotomy (n = 2), papillectomy (n = 1), biliary stent insertion (n = 5), dilation of CBD stenosis with a balloon (n = 4), stone removal (n = 2), and stent retrieval (n = 5)
Moreels et al[38]	Long (EN-450T5, 200 cm)	-	Biliopancreatic pathology (15 procedures), unexplained abdominal pain after gastric bypass (9 procedures), suspected small bowel bleeding or abdominal pain after gastrectomy (4 procedures), enterocutaneous fistula closure (2 procedures)	For biliopancreatic pathology: Cholangiography, balloon dilation, stone extraction, stent placement for unexplained abdominal pain after gastric bypass: Gastric mucosa biopsy. For suspected small bowel bleeding or abdominal pain after gastrectomy: Enteroscopy and mucosal inspection. For enterocutaneous fistula closure: Closure using cyanoacrylate surgical glue and rotatable clip devices
Pohl et al[39]	Long (EN-450T5, 200 cm)	74.6 (49.6-99.6)	Evaluation of jaundice or suspected (n = 14), Biliary obstruction with pain considered to be of biliary origin (n = 1), Biliary obstruction with bleeding (n = 1)	Stone extraction (n = 5), Biliary duct dilation (n = 11), Biliary duct stent placement (n = 3)
Shimatani et al[25]	Short (EC-450BI5, 152 cm)	-	-	Stone extraction (n = 47), Nasobiliary drainage (n = 38), Stent placement (n = 36), Sphincterotomy (n = 31), Choledochojejunostomy dilation (n = 29), Tumor biopsy (n = 10), Nasopancreatic duct drainage (n = 1)
Parlak et al[40]	Long (EN-450T5, 200 cm)	75 (13-137)	Biliary obstruction (n = 11), Cholangitis (n = 9), Stent change (n = 7), Anastomotic stenosis on percutaneous Cholangiogram (n = 1), Jaundice (n = 2), Itching (n = 1)	Stent removal (n = 5), Extraction of catheter (n = 16), Stent placement (n = 10), Stent removal (n = 4), Balloon dilation (n = 1), Stricturoplasty with needle (n = 1)
Cho et al[41]	Short (EC-450BI5, 152 cm)	70.7 (30-117)	Choledocholithiasis (n = 10), Stricture (n = 8), Cholangitis (n = 8), Follow-up ERCP (n = 2), Bile leakage (n = 1)	Sphincterotomy (n = 7), Balloon dilation of biliary sphincter (n = 4), Stone extraction (n = 9), Biliary stricture dilation (n = 8), Stent placement (n = 9), Stent removal (n = 8)
Mönkemüller et al[42]	Long (EN-450T5, 200 cm)	70 (35-240)	Jaundice (n = 3) Cholestasis (n = 7) Choledocolithiasis (n = 4), Cholangitis (n = 3) Upper GI-bleeding (n = 1)	Biliary stent insertion, dilation of common bile duct stenosis with a balloon, stone removal with Dormia basket, stent retrieval
Raithel et al[43]	Long (EN-450T5, 200 cm)	-	Cholestasis, Cholangitis, Choledocholithiasis, Presence of a pancreatic pseudocyst, Suspected or advanced chronic pancreatitis	radiograph the bile ducts (n = 28) Radiography of the pancreatic duct (n = 1)
Chua and Kaffes[44]	Long (EN-450T5, 200 cm)	-	Biliary stenosis (n = 12) Choledocholithiasis (n = 3), Others (n = 1)	Stent removal, stricture dilatation, stone removal with a biliary extraction balloon, plastic stent insertion
Osoegawa et al[45]	Short (EC-450BI5, 152 cm)	93.6 ± 6.8	Pancreatobiliary diseases	Endoscopic biliary drainage (n = 24), Stone extraction (n = 14), Endoscopic sphincterotomy (n = 14), Balloon dilation (n = 10)
Siddiqui et al[46]	Short (EC-450BI5, 152 cm)		Removal of a previously placed stent (n = 5), Sphincter of Oddi dysfunction (n = 3), Surgical biliary leak (n = 3), Pancreatic anastomotic stricture (n = 2), suspected Choledocholithiasis (n = 48), Biliary strictures (n = 18)	Biliary sphincterotomy (n = 39), Dilation of CBD stenosis with a balloon (n = 30), Stent placement (n = 25), Stone extraction (n = 35), Brushing cytology of biliary strictures (n = 3), Stent removal (n = 4)
Shah et al[8]	Short (EC-450BI5, 152 cm); Long (EN-450T5, 200 cm)	(90-120)	Abnormal liver enzyme levels associated with abdominal pain and/or dilated bile ducts on non-invasive imaging (n = 62), Dilated bile ducts on imaging (n = 21), Cholangitis (n = 20), Abnormal liver enzyme levels (n = 11), Pancreatitis (n = 8), Other (n = 7)	Needle-knife precut papillotomy, free-hand (n = 19), Needle-knife precut over a pancreatic stent (n = 2), Needle-knife sphincterotomy over a biliary stent (n = 8), Pull-type sphincterotomy, biliary (n = 12), Pull-type sphincterotomy, pancreatic (n = 3), Anastomotic stricturoplasty in non-transplant hepaticojejunostomy (n = 10), Anastomotic stricturoplasty in transplant hepaticojejunostomy (n = 6), Papillary balloon dilation (n = 12), Stone extraction (1 pancreatic), (n = 21), Non-anastomotic stricture dilation (n = 16), Direct cholangioscopy (n = 11), Biliary stenting (n = 19), Pancreatic stenting (n = 6), Biliary stricture tissue sampling (n = 6)
Varabei et al[47]	Long (EN-450P, 200 cm)	85 (45-120)	Jaundice (n = NR), Cholestasis (n = NR), Cholangiolithiasis (n = NR), Recurrent cholangitis (n = NR), Strictures of HJA (n = NR)	Reconstructive re-operations on the biliary tract (n = 6), Laser vaporizations (n = 3), Stone extraction (n = 2), Lithotripsy (n = 1), Stone extraction and laser vaporization (n = 1)
Cheng et al[26]	Short (EC-450BI5, 152 cm) and Long (EN-450T5, 200 cm)	80 (60-120)	Cholangitis (n = 43), Choledocholithiasis (n = 9), Malignancy (n = 11), Pancreatitis (n = 6), Abnormal liver enzymes Dilated bile duct on imaging (n = 2)	Bile duct stone extraction (n = 49), Cystic duct stone extraction (n = 1), Malignant biliary stricture (n = 6), dilation (n = 9), Biliary dilation alone (n = 9)
Sakakihara et al[48]	Short (EC-450BI5, 152 cm)	-	Stenosis of choledochojejunal anastomosis	Balloon dilation and/or endoscopic biliary stenting
Tsou et al[58]	Short (EC-450BI5, 152 cm); Long (EN-450T5, 200 cm)	45.8 (10-141)	Choledocholithiasis (n = 35), Anastomotic stricture without stones (n = 6), Malignant obstruction (n = 6)	Endoscopic sphincterotomy (n = 4), Balloon dilation (n = 34), Stone extraction (n = 30), Stent placement (n = 10)
Tsutsumi et al[49]	Short (EC-450BI5 and EI-530B, 152 cm)	50 (9-167)		Balloon dilation (n = 47), Stone extraction (n = 17), Endoscopic biliary stenting (n = 18), Endoscopic nasobiliary drainage (n = 28), Peroral direct cholangioscopy (n = 9)
De Koning and Moreels[6]	Long (EN-450T5, 200 cm)	-	Cholangiogram	Biliary anastomosis balloon dilatation Sphincterotomy Precut sphincterotomy Sphincteroplasty Biliary stone removal Biliary plastic stent placement/removal Removal of surgical sutures at biliary anastomosis Treatment of biliary metallic stent ingrowth
Shimatani et al[50]	Short (EI-530B, 152 cm)	22.4 (± 20.8) to reach blind end56.3 (± 32.5) to complete ERC-related interventions	Cholangitis (n = 173), Hepatobiliary disorder without jaundice (n = 60), Obstructive jaundice (n = 43), Bile duct stones (n = 24), Hepatic abscess (n = 4), Others (n = 7)	-
Tomoda et al[59]	Short (EC-450BI5 and EI-530B, 152 cm)	75 (42-180)	Biliary atresia (n = 4), Liver cirrhosis (n = 10), Hepatocellular carcinoma (n = 3), Hepatitis B virus (n = 6), Hepatitis C virus (n = 4), Non-B and non-C hepatitis virus (n = 1), Primary sclerosing cholangitis (n = 1), Primary biliary cirrhosis (n = 2), Alagille syndrome (n = 1), Budd-Chiari syndrome (n = 1)	-
Liu et al[60]	Long (EN-450T5, 200 cm)	-	Biliary obstruction (n = 63), Cholestatic liver function tests (n = 41), Abnormal biliary imaging (n = 31), Ascending cholangitis (n = 14), Stent removal or replacement (n = 17), Bile leak (n = 3), Recurrent pancreatitis (n = 1), Biliary hemorrhage (n = 1), Other (n = 1)	Stent placement (n = 25), Stent removal (n = 11), Dilatation (n = 11), Biopsy/brushing (n = 1)
Shimatani et al[7]	Short (EI-580BT, 155 cm; EI-530B, 152 cm)	54 (37-73)	Bile duct stones (n = 28), Benign stenosis of hepatocyclic acid and hematologic angiogenesis (n = 24), Cholangitis (n = 21), Intrahepatic stones (n = 19), Obstructive jaundice (n = 12), Malignant stenosis of hepatocyclic acid and hematologic angiogenesis (n = 6), Others (n = 2)	Sphincterotomy, Dilation of hepatocyclic acid and hematologic angiogenesis, Stone extraction, Stent placement
Tsutsumi et al[61]	Short (EI-580BT, 155 cm; EI-530B, 152 cm); Long (EN-450T5, 200 cm)	89 (70-103)	Cholangitis (n = 20), Elevated hepatobiliary enzyme (n = 8), Liver abscess (n = 3), Hepatolithiasis (n = 2), Localized intrahepatic bile duct dilation (n = 1)	Stent placement (n = NR), Stone extraction (n = NR)
Kashani et al[31]	Long (450T5, 200 cm)	109 (93-109)	Sphincter of Oddi dysfunction (n = 66), Choledocholithiasis (n = 26), Pancreatitis (n = 9), Biliary stricture/obstruction (n = 8), Bile leak (n = 8), Cholangitis (n = 6), Abnormal liver tests (n = 5), Recurrent liver abscess (n = 1)	-
Matsumoto et al[51]	Short (EI-580BT, 155 cm; EI-530B, 152 cm)	62 (40-74)	Bile leakage (n = 11)	Treatment for bile leakage (drainage) (n = 11)
Mizukawa et al[62]	Short (EI-530B and EC-450BI5, 152 cm)	54 (37-82)	Benign hepatocyclic acid and hematologic angiogenesis structure (n = 52)	Biliary stone extraction (n = 12), Nasobiliary drainage (n = 38)
Yamada et al[63]	Short (EI-580BT, 155 cm; EI-530B, 152 cm)	55 (37-87), EI-580BT; 62 (41-88), EI-530B	Biliary strictures (n = 72), Anastomosis stenoses (n = 113), Choledocholithiasis (n = 94), Intrahepatic stone (n = 55), Miscellaneous (n = 11)	Biliary stone extraction (n = 148), Biliary plastic stenting (n = 101), Biliary metallic stenting (n = 20), Balloon dilation of biliary anastomotic stricture (n = 106), Endoscopic papillary balloon dilation (n = 59), Endoscopic papillary large balloon dilation (n = 24), Endoscopic nasobiliary drainage (n = 151), Pancreatic stone extraction (n = 9), Pancreatic plastic stenting (n = 18), Balloon dilation of pancreatic anastomotic stricture (n = 14), Endoscopic nasopancreatic drainage (n = 16)
Nishio et al[52]	Short (EI-580BT, 155 cm; EI-530B, 152 cm)	43.5 (28.3-61.5) in LO group vs 30 (21-44) in control group	Stenosis of hepaticojejunostomy, bile duct stones, cholangitis, stenosis of biliary duct and others	Balloon dilation, stone extraction, stent placement
Sato et al[53]	Short (EI-580BT, 155 cm; EI-530B and EC-450BI5, 152 cm)	70 (15-240)	Hepaticojejunostomy anastomotic stricture	Balloon dilation and Balloon dilation + plastic stent placement
Sirin and Hulagu[54]	Long (EN-450T5, 200 cm)	-	Biliary stone (n = 15), Stenosis (n = 3), Cholestasis (n = 1), painless jaundice (n = 3), Pancreatic head tumor (n = 3), biliary duct dilatation (n = 3), Cholangiocarcinoma (n = 3)	Extraction of biliary stone by balloon, dilation of stenosis, sphincterotomy and extraction of biliary stone by balloon, biliary stenting, sphincterotomy
Uchida et al[32]	Short (EI-580BT, 155 cm; EI-530B and EC-450BI5, 152 cm)	Median 60	Choledochojejunal anastomosis stenosis (443 procedures), bile duct stones (126 procedures), biliary stricture (177 procedures), pancreatic indications (56 procedures), others (3 procedures)	Plastic stent placement for stenosis, complete stone removal for bile duct stones, diagnostic ERCP
Zamora Nava et al[64]	Long (EN-450T5 and EN-580T, 200 cm); Short (EC-450BI5, 152 cm)	-	Mainly cholangitis (49%), other indications include cholestatic pattern in liver function tests, documented stenosis of hepaticojejunostomy, or lithiasis by imaging studies (CT or MRCP)	Dilation (94.6%), stone extraction (78.6%), sphincterotomy/precut (10.7%), stent placement (7.1%), lithotripsy (1.8%)
CHENG et al[55]	Short (EI-580BT, 155 cm)	115.8 (40-275)	Bile duct stone (40.5%), painless obstructive jaundice (40.5%), acute cholangitis (13.5%), acute pancreatitis (5.5%)	Stone extraction or mechanical lithotripsy (n = 23), Sphincterotomy/sphincteroplasty (n = 18), Stricture dilatation (n = 7), Plastic stent insertion (n = 11), Metallic stent insertion (n = 6), Biopsy/brush cytology (n = 4)
Obata et al[56]	Short (EI-580BT, 155 cm; EI-530B, 152 cm)	90 (67-120)	CBD stones	Precut alone 4 (n = 6), EST alone 6 (n = 10), EPBD alone 9 (n = 14), EPLBD alone 5 (n = 8), Precut + EPBD 12 (n = 19), Precut + EPLBD 5 (n = 8), EST + EPBD 12 (n = 19), EST + EPLBD 5 (n = 8), None 5 (n = 8)
Sato et al[57]	Short (EI-580BT, 155 cm)	63 (35-86)	Treatment of hepaticojejunostomy anastomotic strictures	Placement of fully-covered metal stent (FCSEMS) for HJASs; balloon dilation of stricture prior to stent placement; additional plastic stent placement in some cases as “rescue stent”
Farina et al[65]	Short (EI-580BT, 155 cm); Long (EN580T, 200 cm)	-	Cholangitis or jaundice caused by benign/malignant biliary stenosis of the common bile duct or stenosis of a biliodigestive anastomosis, common bile duct gallstones, biliary leak, and removal of previously placed stents	Sphincterotomy, stone extraction, or pancreaticobiliary stent placement
Yokoyama et al[66]	Short (EI-580BT, 155 cm; EI-530B and EC-450BI5, 152 cm); Long (EN-450P5 and EN-450T5, 200 cm)	96 ± 43	Biliary stone	Stone extraction (n = 257)

CBD: Common bile duct; ERCP: Endoscopic retrograde cholangiopancreatography; LFT: Liver function tests; CT: Computed tomography; ERC: Endoscopic retrograde cholangiography; EST: Endoscopic sphincterotomy; EPBD: Endoscopic papillary balloon dilation; EPLBD: Endoscopic papillary large balloon dilation.

DBEERCP outcomes

Table 5 summarizes the DBE-ERCP outcomes for patients with SAA. Among the included studies, 87.5% (35/40) documented papilla reach, 75% (30/40) reported enteroscopic success, 65% (26/40) included diagnostic success rates, procedural success was described in 97.5% (39/40), and 87.5% (35/40) provided information on adverse events following the procedure.

Table 5 Double-balloon enteroscopy-assisted endoscopic retrograde cholangiopancreatography outcomes.

Ref.	Anastomosis/papilla reached, % (n/N)	Enteroscopy success, % (n/N)	Diagnostic success, % (n/N)	Procedural success, % (n/N)	Adverse events, % (n/N)	Adverse event description
Aabakken et al[34]	94 (17/18)	83.3 (15/18)	100 (15/15)	100 (9/9)	0 (0/16)	-
Emmett and Mallat[35]	85 (17/20)	94.1 (16/17)	-	100 (10/10)	0 (0/20)	-
Maaser et al[36]	72.7 (8/11)	64 (7/11)	64 (7/11)	100 (5/5)	0 (0/11)	-
Mönkemüller et al[37]	90.9 (10/11)	90.9 (10/11)	81.81 (9/11)	60 (6/10)	9.09 (1/11)	Perforation of hepatocolejunostomy during balloon dilatation (n = 1)
Moreels et al[38]	86.7 (13/15)	80 (12/15)	80 (12/15)	80 (12/15)	13.3 (2/15)	Perforation (n = 2)
Pohl et al[39]	84 (22/25)	84 (22/25)	-	100 (16/16)	0 (0/15)	-
Shimatani et al[25]	97 (100/103)	-	98 (98/100)	100 (98/98)	5 (5/103)	Retroperitoneal emphysema (n = 1), Intestinal perforation (n = 1), Retroperitoneal perforation (n = 1), Post-endoscopic sphincterotomy perforation (n = 1), Subcutaneous emphysema with pneumothorax (n = 1)
Parlak et al[40]	92.86 (13/14)	92.86 (13/14)	92.86 (13/14)	85.71 (12/14)	7.14 (1/14)	Perforation (n = 1)
Cho et al[41]	86.2 (25/29)	82.76 (24/29)	83 (16/20)	100 (24/24)	0 (0/20)	-
Mönkemüller et al[42]	-	77.8 (14/18)	92.9 (13/14)	69.2 (9/13)	5.6 (1/18)	Perforation (n = 1)
Raithel et al[43]	74.2 (23/31)	-	-	87.2 (75/86)	5.8 (5/86)	Post-ERCP pancreatitis (n = 2), Post-interventional bleeding (n = 1), Perforation (n = 2)
Chua and Kaffes[44]	-	77.8 (14/18)	77.8 (14/18)	80 (8/10)	-	-
Osoegawa et al[45]	95.7 (45/47)	88.9 (40/45)	88.9 (40/45)	100 (40/40)	2.1 (1/47)	Perforation (n = 1)
Siddiqui et al[46]	89.9 (71/79)	90.1 (64/71)	90.1 (64/71)	100 (64/64)	5.1 (4/79)	Mild post-ERCP pancreatitis (n = 3), Self-limited bleeding (n = 1)
Shah et al[8]	74 (20/27)	-	-	63 (17/27)	-	Pancreatitis (n = 5), Mild bleeding (n = 1), Abdominal pain requiring hospital admit (n = 3), Throat pain requiring physician contact (n = 4), Perforation (n = 2), Death (n = 1), Reported for all interventions including SBE, DBE, and rotational ERCP
Varabei et al[47]	63.6 (21/33)	63.6 (21/33)	61.9 (20/33)	57.14 (12/21)	-	-
Cheng et al[26]	95 (73/77)	89.61 (69/77)	-	87 (67/77)	6.5 (5/77)	Minor perforation (n = 2), Mucosal tear (n = 2), Intestinal perforation (n = 1)
Sakakihara et al[48]	86.4 (38/44)	-	-	81.8 (36/44)	15.9 (7/44)	Cholangitis (n = 7)
Tsou et al[58]	83 (39/47)	94.9 (37/39)	94.9 (37/39)	100 (37/37)	4.3 (2/47)	Intestinal perforation (n = 1), Biliary tract infection (n = 1)
Tsutsumi et al[49]	98.61 (71/72)	98.61 (71/72)	98.61 (71/72)	100 (59/59)	2.72 (2/72)	Cholangitis (n = 2)
De Koning and Moreels[6]	-	-	-	73 (22/30)	10 (3/30)	Retroperitoneal free air (n = 3), Post-ERCP cholangitis (n = 2), Post-ERCP pancreatitis (n = 2), Liver capsule dehiscence (n = 1), (List includes complications for both DBE and SBE procedures in study)
Shimatani et al[50]	97.7 (304/311)	-	96.4 (293/304)	97.9 (277/283)	10.6 (33/311)	Obvious perforation (n = 1), Microperforation (n = 6), Mucosal laceration (n = 1), Biliary damage (n = 4), Pancreatitis (n = 11), Cholangitis (n = 8), Aspiration pneumonia (n = 2)
Tomoda et al[59]	85 (17/20)	85 (17/20)	70 (14/20)	55 (11/20)	10 (2/20)	Bleeding (n = 1), Ischemic liver graft (n = 1)
Liu et al[60]	76 (65/86)	70 (60/86)	100 (1/1)	89.4 (42/47)	2.33 (2/86)	Post-procedure cholangitis (n = 1), Aspiration pneumonia (n = 1)
Shimatani et al[7]	99.1 (111/112)	-	98.2 (109/111)	100 (109/109)	2.7 (3/112)	Laceration in intestinal tract (n = 3)
Tsutsumi et al[61]	97 (32/33; 30/33 short DBE, 2/33 Long DBE)	97 (32/33; 30/33 short DBE, 2/33 Long DBE)	81 (26/32; 21/32 short DBE, 5/32 Long DBE)	-	18.18 (6/33)	Perforation and pneumatosis (n = 3), Cholangitis (n = 1), Elevated hepatobiliary enzymes (n = 1), Mucosal laceration of ilium (n = 1)
Kashani et al[31]	93.8 (121/129)	94.57 (87/92)	-	94.57 (87/92)	10.1 (13/129)	Pancreatitis (n = 10), Cholangitis (n = 1), Perforation (n = 2)
Matsumoto et al[51]	100 (11/11)	100 (11/11)	-	63.6 (7/11)	0 (0/11)	-
Mizukawa et al[62]	100 (46/46)	100 (46/46)	100 (46/46)	100 (46/46)	6.52 (3/46)	Cholangitis (n = 3)
Yamada et al[63]	EI-580BT: 100 (163/163) EI-530B: 100 (163/163) Cumulative: 100 (326/326)	EI-580BT: 96 EI-530B: 87 Cumulative: 91.5	-	EI-580BT: 92 (150/163) EI-530B: 89 (145/163) Cumulative: 90.5 (295/326)	EI-580BT: 4.3 (7/163) EI-530B: 6.7 (11/163) Cumulative: 5.5 (18/326)	Pancreatitis (n = 11), Perforation (n = 7)
Nishio et al[52]	95.2 (99/104)	84.5 (81/97)	93.8 (91/97)	97.4 (75/77)	8.7 (9/104)	Cholangitis (n = 5), pancreatitis (n = 2), bleeding (n = 2)
Sato et al[53]	-	-	-	89.2 (91/102)	17.6 (18/102)	Cholangitis (n = 10), Intra-abdominal/retroperitoneal air (n = 6), Bile leakage (n = 1), Bleeding (n = 1)
Sirin and Hulagu[54]	96.8 (30/31)	96.8 (30/31)	87.1 (27/31)	87.1 (27/31)	6 (2/31)	Perforation (n = 1), Retroperitoneal abscess (n = 1)
Uchida et al[32]	94.3 (759/805)	-	-	90.7 (730/805)	5.5 (44/805)	Cholangitis (n = 20) pancreatitis (n = 14), intestinal perforation (n = 6), others (4 cases, including 2 biliary leakage, 1 pancreatic leakage caused by guidewires, and 1 pneumothorax)
Zamora Nava et al[64]	-	77.1 (74/96)	69.8 (67/96)	83.6 (56/67)	14.5 (14/96)	Cholangitis (n = 7), Pancreatitis (n = 2), Perforation (n = 2), Mucosal tear (n = 2), Hemorrhage (n = 1)
CHENG et al[55]	95.7 (44/46)	-	84.1 (37/44)	76.1 (35/46)	13 (6/46)	Intestinal perforation (n = 2), acute cholangitis (n = 3), acute pancreatitis (n = 1)
Obata et al[56]	92.4 (73/79)	81 (64/79)	80 (63/79)	78 (62/79)	5 (4/79)	Bowel perforation (n = 2), pancreatitis (n = 1), hypoxia (n = 1)
Sato et al[57]	100 (20/20)	100 (20/20)	100 (20/20)	100 (20/20)	-	-
Farina et al[65]	69.8 (37/53)	31/37 (83.8)	-	31/37 (84)	0 (0/53)	-
Yokoyama et al[66]	91.3 (264/289)	83.4 (241/289)	-	100 (241/241)	-	-

ERCP: Endoscopic retrograde cholangiopancreatography; DBE: Double-balloon enteroscopy.

The results of the pooled proportion meta-analysis for all outcomes are presented as follows. The combined success rate for reaching the papilla was 92% (95%CI: 89%-95%). Overall success rate of enteroscopy was 89% (95%CI: 85%-92%). The pooled estimate of diagnostic success rate was 90% (95%CI: 85%-95%). The procedural success rate was 92% (95%CI: 89%-95%) (Supplementary Figures 1, 2, 3, and 4). Heterogeneity was high across all outcomes, ranging from I² of 73.3% to 87.8%. Publication bias was assessed using funnel plots (Figure 2A-D) and Egger’s regression test. The test indicated significant asymmetry for papilla reached (intercept = 3.12, SE = 1.07, t = 2.93, df = 18, P = 0.009), enteroscopy success (intercept = 1.64, SE = 0.75, t = 2.19, df = 28, P = 0.037), and diagnostic success (intercept = 3.11, SE = 1.17, t = 2.67, df = 24, P = 0.014), suggesting potential publication bias and possible overestimation of effect size in smaller studies. No significant asymmetry was observed for procedural success (intercept = 0.50, SE = 0.54, t = 0.92, df = 37, P = 0.37), indicating no evidence of publication bias for this outcome. To account for potential bias, a sensitivity analysis excluded studies including patients with Billroth II reconstruction, who tend to have shorter biliopancreatic limbs and higher success rates. Across all measured outcomes, results were similar: 92% (95%CI: 89%-95%) for reaching the papilla, 89% (95%CI: 84%-94%) for enteroscopic success, 92% (95%CI: 83%-98%) for diagnostic success, and 88% (95%CI: 79%-94%) for procedural success.

Figure 2 Funnel plot and Egger’s regression. A: Success rate for reaching the papilla; B: Success rate of enteroscopy; C: Diagnostic success rate; D: Procedural success rate; E: Incidence of complications.

After excluding the 7 studies that used both short-scope and long-scope DBE, we conducted a subgroup analysis between the two types using the remaining 33 studies. The results are summarized in Table 6. The P values for papilla reached rate, enteroscopy success rate, and procedural success rate were all below 0.05, indicating statistical significance, whereas those for diagnostic success rate and adverse event rate were above 0.05, suggesting no significant difference.

Table 6 Summary of proportion meta-analysis of all outcomes (after excluding the 7 studies that used both short-scope and long-scope double-balloon enteroscopy).

Type of DBE	Papilla reached (%)	Enteroscopy success (%)	Diagnostic success (%)	Procedural success (%)	Adverse events (%)
Overall	94 (95%CI: 90-96), I2 = 83.6	89 (95%CI: 84-93), I2 = 74.9	92 (95%CI: 86-96), I2 = 80.9	93 (95%CI: 88-96), I2 = 85.5	5.5 (95%CI: 4-7.2), I2 = 45.0
Short DBE	97 (95%CI: 94-99), I2 = 79.6	92 (95%CI: 86-97), I2 = 74.0	94 (95%CI: 88-98), I2 = 80.2	89 (95%CI: 89-99), I2 = 89.6	6.2 (95%CI: 4.1-8.6), I2 = 59.9
Long DBE	86 (95%CI: 78-92), I2 = 67.2	84 (95%CI: 76-91), I2 = 69.5	87 (95%CI: 74-96), I2 = 80.4	87 (95%CI: 79-94), I2 = 62.5	4.5 (95%CI: 2.2-7.4), I2 = 6.7
Test for subgroup differences P value	P < 0.0001	P = 0.0289	P = 0.1249	P = 0.0275	P = 0.7069

DBE: Double-balloon enteroscopy.

Adverse events

Table 5 provides a comprehensive overview of adverse events following DBE-ERCP. The overall pooled incidence of complications was 5.7% (95%CI: 4.1%-7.5%) (Supplementary Figure 5). A total of 221 patients experienced DBE-ERCP-related complications, with the most frequently reported being cholangitis (n = 70), pancreatitis (n = 59), and perforation (n = 47). One patient developed an intestinal perforation during endoscope insertion and required emergency laparotomy. No deaths attributable to DBE-ERCP were reported. Considerable heterogeneity was observed across the studies (I² = 53.3%; P = 0.0001). Publication bias was assessed using funnel plots and Egger’s regression test (Figure 2E). For adverse events, Egger’s test showed no significant asymmetry (intercept = -0.68, SE = 0.40; t = -1.70, df = 33; P = 0.10), indicating no evidence of publication bias. A sensitivity analysis was performed by excluding studies involving patients with Billroth II reconstruction, yielding an overall rate of 5% (95%CI: 3%-7%).

DISCUSSION

Our study indicates that, based on current evidence, this technique effectively enables ERCP in patients with diverse surgical limb lengths-from short Billroth II to Roux-en-Y gastric bypass (RYGB)-while showing a low incidence of adverse events. The 18 Japanese studies included in this meta-analysis provided the majority of the data, covering all anatomical types. Due to the low prevalence of severe morbid obesity in Japan, RYGB is not commonly performed for obesity. In contrast, RYGB is a standard treatment for morbid obesity in the United States. All four United States studies included in our analysis focused on RYGB procedures. Most of the five German studies included in this meta-analysis focused on Roux-en-Y anatomy. The global obesity epidemic continues to grow at an alarming rate, affecting more than 2 billion people worldwide[19]. Patients undergoing bariatric surgery often develop biliary diseases associated with weight loss. As a result, an increasing number of patients may require ERCP through the Roux-en-Y limb in the future. For patients with Billroth II anatomy, various endoscopes, such as duodenoscopes and DAE, achieve comparable technical success rates. Accordingly, ESGE guidelines recommend duodenoscopy as the initial endoscopic modality of choice[20]. The inclusion of patients with Billroth II reconstruction may have contributed to the overall high success rate. However, even after excluding studies involving these patients, the results remained comparable to those of the overall analysis.

Patients who have undergone Roux-en-Y or Whipple procedures often present with longer and/or more sharply angulated afferent limbs, which can significantly reduce the likelihood of successful cannulation and therapeutic success. In patients with altered anatomy, there are typically two main approaches to access the biliopancreatic system. Transluminal access is achieved using DAE, while transmural access is established through therapeutic endoscopic ultrasound (EUS) or surgical intervention. Recently developed techniques, such as EUS-guided antegrade for the management of bile duct stones and EUS-guided biliary drainage (EUS-BD) for biliary decompression, have demonstrated technical success rates comparable to those of device-assisted ERCP[21,22]. Nonetheless, as these techniques are still in the early stages of clinical application, further standardization and protocol optimization are required to support their broader adoption. Multicenter randomized controlled trials are likely required to assess the comparative risks and benefits of various techniques and to determine the optimal approach for enteroscopy and ERCP in this patient group.

DBE, originally developed to enable deep examination of the small intestine, was introduced by Yamamoto et al[23] in 2001. The use of DBE in patients with SAA allows access to the bilioenteric anastomosis or papilla through long intestinal limbs, significantly increasing the success rate of ERCP in this patient population. Our meta-analysis of the published literature indicates that DBE-ERCP is highly effective, achieving diagnostic and therapeutic success rates of 90% and 92%, respectively, with an overall adverse event rate of 5.7%. When excluding Billroth II cases, the diagnostic and procedural success rates were 92% and 88%, respectively, with a 5% incidence of complications. Currently, two DBE systems are available: The conventional long-scope (200 cm scope) and a newer short-scope (< 200 cm). The newer short-scope system is more convenient and compatible with conventional ERCP accessories[24,25]. Short-scope DBE is preferentially used in patients with short afferent biliopancreatic limbs (< 50 cm), such as those who have undergone Billroth II or shorter Roux-en-Y reconstructions-including Roux-en-Y with intact papilla or Roux-en-Y with bilioenteric and/or pancreatoenteric anastomosis. Conversely, ERCP using a long-scope DBE may be better suited for anatomical configurations with limbs longer than 100 cm, like RYGB[3]. Previous studies have confirmed that both approaches have comparable success and safety profiles, consistent with the findings from the subgroup analysis of our data[8,26]. SBE and manual spiral enteroscopy later emerged as potential alternatives, with comparable success rates and safety profiles[27,28]. A recent systematic review and meta-analysis on balloon-assisted and spiral enteroscopy-assisted ERCP also demonstrated this[29]. Manual spiral enteroscopy was subsequently replaced by the motorized spiral enteroscope (MSE) in 2016[30]. Although MSE-assisted ERCP shows technical potential in patients with SAA, its current design limits procedural efficiency and safety. Compared with BAE or EUS-BD, MSE has not demonstrated the expected advantages. The previous generation double-balloon endoscope, the EI-530B (FUJIFILM Co., Tokyo, Japan), features a 1520 mm working length and a 2.8 mm working channel. The latest short-scope DBE model, the EI-580BT (FUJIFILM Co., Tokyo, Japan), features a working length of 1550 mm and an expanded 3.2 mm channel. The most notable advancement of the new model is its enlarged working channel, which, combined with advanced force transmission and an adaptive bending system, enables the use of a wider range of instruments and significantly enhances procedural flexibility[7]. Current evidence suggests that with increasing operator experience, the time required for papilla cannulation during DBE-ERCP gradually decreases; however, the overall cannulation success rate does not change significantly[31]. DBE-ERCP, as a resource-intensive technique, faces certain limitations in terms of cost and availability. From a cost perspective, DBE systems comprise specialized enteroscopes, overtubes, and balloon inflation devices, resulting in relatively high equipment acquisition expenses. Roux-en-Y reconstruction and first-time short DBE-ERCP were factors affecting the technical failure and adverse event rates. Trainees should observe multiple DBE-ERCP procedures performed by experts to learn the techniques or focus on gaining experience with initial procedures or relatively simple cases[32]. A more advanced instrument and the accumulation of operator experience are key to improving the procedural success rates and reducing adverse events in DBE-ERCP. In terms of availability, DBE equipment may not be readily accessible in many healthcare centers, particularly in resource-limited regions, which restricts the widespread implementation of this technique. Furthermore, although DBE-ERCP is highly effective in certain scenarios, combining it with EUS-guided interventions could provide patients with additional safety and therapeutic advantages[33].

This study integrates data from 40 cohort studies or case series, making it the largest meta-analysis to date. However, several limitations should be noted. Most of the studies included involved diverse patient populations (in terms of anatomical structure, interventions, and patient characteristics), resulting in significant heterogeneity in surgical success rates. The majority of the included studies (34 out of 40) were retrospective, introducing inherent bias. Most of the studies were small in scale and observational in design, utilizing single-arm, non-randomized approaches, which limits the overall quality of the evidence. Although this study demonstrates the overall effectiveness of DBE-ERCP, the lack of subgroup-specific data prevents a reliable assessment of success rates across different anatomical configurations. Future studies with more detailed subgroup analyses may offer more tailored clinical guidance. However, despite the heterogeneity and potential variations in success rates related to these factors, our overall findings indicate that DBE-ERCP remains effective and demonstrates broad applicability in clinical practice. In some studies, patients initially underwent treatment attempts using standard colonoscopes or other endoscopic devices, which were unsuccessful, before DBE was employed. As a result, the technical difficulty for this patient group was inherently higher. Some studies report the success rate as the ratio of successful procedures to the total number of DBE-ERCP attempts-including multiple ERCPs per patient-whereas most studies report it as the proportion of patients in whom the procedure was ultimately successful out of the total patient population. The time interval between surgery and DBE-ERCP may affect procedural success and surgical safety. However, most of the included studies did not report this information. Future studies should be more comprehensive and in-depth to address these gaps and provide more precise guidance for clinical practice.

CONCLUSION

Current clinical evidence suggests that DBE-ERCP is both feasible and safe for the diagnosis and treatment of pancreatobiliary diseases in patients with SAA. The advantages of the short-scope double-balloon enteroscope as a newer device were partially confirmed in this study. Further randomized controlled trials are needed to strengthen the reliability of these findings.