Abstract
We aimed to compare enteroscopy-assisted ERCP (EA-ERCP), laparoscopy-assisted ERCP (LA-ERCP), and endoscopic ultrasound-directed ERCP (EDGE) in terms of safety and efficacy in post-Roux-en-Y gastric bypass patients. We conducted a rigorous analysis based on a predefined protocol (PROSPERO, CRD42022368788). Sixty-seven studies were included. The technical success rates were 77% (CI 69–83%) for EA-ERCP, 93% (CI 91–96%) for LA-ERCP, and 96% (CI 92–98%) for EDGE. Subgroup differences were significant between the EA-ERCP and other groups (p < 0.05). The overall adverse event rates were 13% (CI 8–22%), 19% (CI 14–24%), and 20% (CI 12–31%), respectively (p = 0.49). Our findings suggest that EDGE and LA-ERCP may be more effective and as safe as EA-ERCP.
Graphical abstract
Supplementary Information
The online version contains supplementary material available at 10.1007/s11695-024-07459-z.
Keywords: ERCP, Surgically altered anatomy, Laparoscopy, Enteroscopy, Endoscopic ultrasound
Introduction
The ever-growing burden of the obesity pandemic has increased the number of bariatric surgeries performed. An epidemiological survey showed that almost 49,000 Roux-en-Y gastric bypass (RYGB) operations were performed in 26 European countries in 2018, making it the second most common bariatric surgery type [1]. A patient with an RYGB anatomy admitted for pancreaticobiliary intervention raises difficult questions for endoscopists regarding the access routes to the papilla of Vater.
Endoscopic retrograde cholangiopancreatography (ERCP) is often required in these patients, primarily because of the high incidence of gallstone formation after rapid weight loss, but other indications are also common. The two current gold standard ERCP techniques are enteroscopy-assisted ERCP (EA-ERCP) [2] and laparoscopy-assisted ERCP (LA-ERCP) [3, 4], and the most recent one is endoscopic ultrasound-directed transgastric ERCP (EDGE) [5, 6]. Because the three modalities require different technical skills, expertise, and technological backgrounds, research has been conducted to determine the advantages and disadvantages of each procedure. Recent data suggest that EDGE and LA-ERCP appear more effective but less safe than EA-ERCP [7–12]. Cost-effectiveness and cost-utility analyses found EDGE the best option, and LA-ERCP was the costliest [13, 14]. This controversy has created a dilemma in therapeutic decision-making.
The current study aimed to address this issue by conducting a systematic review and meta-analysis to investigate the safety and efficacy of EA-ERCP, LA-ERCP, and EDGE in patients who previously underwent RYGB.
Materials and Methods
This study was performed according to a predetermined protocol registered on the PROSPERO website (CRD42022368788), and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist was used to design and conduct the study.
Eligibility Criteria
The PICO framework was used to address the questions posed in this systematic review. Studies reporting consecutive patients with RYGB anatomy (P) undergoing EDGE and/or LA-ERCP and/or EA-ERCP (I/C) and reporting technical and clinical success and/or complications (O) were included. Both interventional and observational studies were eligible for inclusion. Comparative (multi-arm) and non-comparative (one-arm) studies were also included. Case reports and congress abstracts were excluded from the study. Detailed information on the inclusion criteria is provided in Sect. "Eligibility Criteria" of Supplementum.
Systematic Search
A comprehensive search was conducted in three databases: MEDLINE (via PubMed), Embase, and the Cochrane Central Register of Controlled Trials (CENTRAL) Library on November 20, 2022. A backward and forward reference search of eligible studies was performed on January 12, 2023, using The Lens citation chaser tool. More information on the systematic search and detailed search key can be found in Document 1 of Supplementum.
Selection Process
Information on the selection process is available in Supplementum, Sect. "Selection Process".
Data Collection
A description of the data collection process can be found in Supplementum, Sect. "Data Collection".
Data Items
A description of the collected data item can be found in Supplementum, Sect. "DataItems".
Outcomes
Primary outcomes included technical success rate and clinical success rate.
The secondary outcomes were the number of overall (pooled) and individual adverse events. More information on the outcomes can be found in Supplementum, Sect. "Outcomes".
Effect Measures
Proportions with a 95% confidence interval (CI) for indirect comparisons and odds ratios (OR) with 95% CI for direct comparisons were used for the effect size measure. Detailed information on the effect measures is provided in Supplementum, Sect. "Effect Measures".
Synthesis Methods
For a detailed description of the statistical synthesis methods, see Supplementum, Sect. "Synthesis Methods".
Risk of bias (RoB)
The RoB evaluation was performed independently by two reviewers (B.G. and J.H.). Disagreements were resolved by a third independent reviewer (A.R.). The Risk of Bias in Non-randomized Studies (ROBINS-I) tool was used for the observational interventional studies. RoB for non-comparative studies was assessed using the Methodological Index for Non-Randomized Studies (MINORS) tool. In this review, a MINORS score ≤ 8 was considered high risk, 9–14 moderate risk, and 15–16 low risk of bias. The RoB assessment tools are described in Supplementum, Sect. "Risk of bias (RoB)".
Certainty Assessment
Quality of evidence (QoE) was assessed using the GRADE tool. Two reviewers (B.G. and J.H.) conducted the GRADE assessment independently. Disagreements were resolved by a third independent reviewer (A.R.). The GRADE tool is described in the Supplementum, Sect. "Certainty Assessment".
Results
Search and Selection
Altogether, 8882 hits were found after duplicate removal, and 2714 patients from 67 studies were included in the review (Fig. 1).
Fig. 1.
PRISMA 2020 flowchart of article selection process
Baseline Characteristics
The baseline characteristics of the included studies are shown in Table 1. Of these, 62 were retrospective, and 5 were prospective in design. Seven studies compared two different interventions. A total of 35 studies reported on LA-ERCP, 13 on EDGE, and 12 on EA-ERCP alone.
Table 1.
Baseline characteristics of included studies
| Number | Author, year | Study design | Study period, country | Age (years)1 | Sex (female% of total) | Indications (for every ERCP procedures of interest if it is possible, if not than for the whole study) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CDL (n) | Cholangitis (n) | Malignant stenosis (n) | Benign stenosis (n) | Undefined biliary stricture/obstruction (n) | Acute pancreatitis (n) | SOD/papillary stenosis (n) | Abnormal LFTs/abnormal imaging (n) | Bile leak (n) | Other (n) | ||||||
| Studies about EDGE | |||||||||||||||
| 1 | Chhabra, 2022 | retrospective single-arm interventional | January 2020–October 2021, UK | 65.8 ± 9.8 | 85.7 | 13 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0 |
| 2 | Tyberg, 2020** | retrospective single-arm interventional | over 3 years, not specified, USA | 58.7 ± 10.6 | 79 | 5 | 0 | 0 | 0 | 6 | 3 | 0 | 0 | 0 | 5 |
| 3 | Krafft, 2021 | Retrospective cohort | March 2018–February 2020, USA | SS-EDGE: 54.9 ± 11.9; DS-EDGE: 60.5 ± 10.8 | 66.7 | 12 | 2 | 4 | 0 | 0 | 1 | 0 | 0 | 2 | 0 |
| 4 | Krafft 2022 | prospective cohort | March 2018–October 2019, USA | 62.4 ± 9.5 | 81.8 | 13 | 0 | 0 | 3 | 0 | 0 | 1 | 0 | 2 | 3 (1 IPMN, 1 CP, benign gastric remnant wall thickening on CT) |
| 5 | Bahdi 20222 | retrospective single-arm interventional | February 2018–November 2019, USA | 57 (IQR 51–62.25) | 89.7 | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
| 6 | Runge, 20212 | retrospective single-arm interventional | February 2015–March 2019, USA, UK | 58 ± 11 | 79 | 97 | 0 | 8 | 16 | 0 | 0 | 0 | 8 | 18 | 19 (1 CP, 2 cholangioscopy, 4 PD stone, 12 other incl. Foreign body and SOD) |
| 7 | Shinn, 2021 | retrospective single-arm interventional | March 2016–October 2019, USA | Median: 58 | 77 | 0 | 13 | 0 | 0 | 91 | 0 | 0 | 0 | 15 | 0 |
| 8 | Ghandour, 2022 | retrospective case–control | June 2015–September 2021, USA, UK | 57.9 ± 11.1 | 88 | 38 | 0 | 3 | 10 | 0 | 7 | 7 | 0 | 6 | 0 |
| 9 | Keane, 2022** | retrospective single-arm interventional | February 2018–May 2021, USA | 58.1 ± 11.1 | 86.5 | 16 | 0 | 7 | 0 | 3 | 3 | 4 | 0 | 4 | 0 |
| 10 | Barclay, 2022 | retrospective single-arm interventional | January 2019–December 2021, Canada | Median: 67 (40–72) | 57.1 | 6 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| 11 | James, 2019** | retrospective single-arm interventional | January 2016–January 2018, USA | 55.5 ± 3.2 | 78.9 | 8 | 0 | 0 | 3 | 0 | 6 | 1 | 1 | 0 | 0 |
| 12 | Sanz, 2020 | cases series | October 2016–July 2019, Spain | 56 ± 9.7 | 85.7 | 6 | 3 | 0 | 0 | 0 | 2 | 0 | 1 | 1 | 1 |
| 13 | Irani, 2018 | cases series | June 2015–August 2017, USA | 52 (31–71) | 60 | 3 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Studies about EA-ERCP | |||||||||||||||
| 14 | Dellon, 2009*, ** | retrospective single-arm interventional | January 2008–May 2008 | USA | 1 | 50.5 | 75 | 1 | 4 | 4 | 100 | 0 | 0 | 0 | 0 |
| 15 | Lennon, 2012** | retrospective single-arm interventional | October 2007–March 2011 | USA | 1 | 52.8 ± 14.4 | 76.5 | 12 | 34 | 54 | n/a | n/a | n/a | n/a | n/a |
| 16 | Siddiqu, 2012i** | retrospective single-arm interventional | April 2008–November 2011 | USA | 2 | 58 (29–86) | 62 | 39 | 79 | 79 | n/a | 48 | 0 | 0 | 0 |
| 17 | Saleem, 2010** | retrospective single-arm interventional | April 2008–February 2010 | USA | 1 | 57 (19–85) | 32 | 15 (cases not patients) | 50 | 56 | n/a | n/a | n/a | n/a | n/a |
| 18 | Trindade, 2015** | retrospective single-arm interventional | July 2011–September 2013 | USA | 1 | 56 (28–80) | 80.4 | 44 | n/a | n/a | 100 | 29 | 10 | 0 | 0 |
| 19 | Elton, 1998** | retrospective single-arm interventional | March 1994–June 1997 | USA | 1 | 53 (24–76) | 55.6 | 3 | 18 | 18 | n/a | 0 | 4 | 4 | 0 |
| 20 | Emmett, 2007 | cases eries | October 2005–January 2007 | USA | 1 | 50 (40–73 | 50 | 6 | 14 | 20 | n/a | 0 | 2 | 0 | 0 |
| 21 | Wagh, 2012** | prospective cohort | January 2009–June 2011 | USA | 1 | 59.3 (39–72) | 86 | 4 | 7 | 13 | n/a | 1 | 0 | 0 | 0 |
| 22 | Ali, 2018**2 | retrospective single-arm interventional | December 2009–October 2016 | USA | 1 | Median: 55 (22–75) | 80.6 | 28 (cases) | 31 | 35 | n/a | 14 | 0 | 6 | 1 |
| 23 | Choi, 20132 | retrospective single-arm interventional | December 2005–November 2011 | USA | 1 | 56.1 ± 12.2 | 92.9 | 28 | 72 | 32/108 (DBE/total) | 79 | 16 | 0 | 0 | 0 |
| 24 | Sawas, 2019** | retrospective single-arm interventional | January 2015–December 2017 | USA | 1 | 62.2 ± 10.2 | 77.8 | 18 | 30 | 20/62 (BAE/total) | 83.3 | 20 | 0 | 2 | 7 |
| 25 | Kashani, 20182 | retrospective cohort | December 2005–July 2012 | USA | 1 | Median: 50 (22–82) | 87.4 | 103 | 103 | 129 | n/a | 26 | 6 | 0 | 0 |
| Studies about LA-ERCP | |||||||||||||||
| 26 | Brockmeyer, 2015 | retrospective single-arm interventional | September 2001–September 2014 | USA | 1 | n/a | n/a | 8 | n/a | n/a | 100 | 8 | 0 | 0 | 0 |
| 27 | Patel, 2008* | retrospective single-arm interventional | January 2007–December 2007 | USA | 1 | 46.5 ± 11.2 | 83 | 6 | 8 | 8 | 100 | 6 | 0 | 0 | 0 |
| 28 | Richardson, 2012* | retrospective single-arm interventional | January 2006–December 2011 | USA | 1 | 54.3 ± 12.6 | 81.8 | 11 | 13 | 13 | 91 | 4 | 0 | 0 | 0 |
| 29 | Baimas-George, 2020 | retrospective single-arm interventional | January 2010–December 2017 | USA | 1 | 54 (30–69) | 82.5 | 40 | 40 | 40 | 100 | 21 | 8 | 0 | 0 |
| 30 | Ivano, 2019* | retrospective single-arm interventional | January 2007–January 2017 | Brasil | 1 | 50 ± 14.6 | 57 | 7 | 7 | 7 | 100 | 7 | 0 | 0 | 0 |
| 31 | Bayoumi, 2016 | prospective cohort | January 2013–June 2015 | Egypt, Saudi Arabia | 2 | 44.8 ± 5.8 | 75 | 12 | 12 | 12 | 100 | 6 | 0 | 0 | 0 |
| 32 | Snauwaert, 2015 | retrospective single-arm interventional | May 2008–September 2014 | Belgium | 5 | Median: 54 (26–79) | 78.3 | 23 | 23 | 23 | 95.7 | 16 | 0 | 0 | 0 |
| 33 | May, 2019 | retrospective single-arm interventional | January 2009–December 2016 | USA | 1 | 55.4 ± 10.9 | 88 | 51 | 51 | 51 | 57 | 24 | 0 | 0 | 0 |
| 34 | Paranandi, 2016* | retrospective single-arm interventional | September 2011–June 2013 | UK | 1 | 50.3 ± 14.4 | 100 | 7 | 7 | 7 | 71.4 | 4 | 1 | 0 | 0 |
| 35 | Popowicz, 2021** | retrospective single-arm interventional | January 2007–December 2013 | Sweden | n/a | 42 ± 14 | 93 | 2 | 51 | 51 | 100 | 2 | 0 | 0 | 0 |
| 36 | Gutierrez, 2009* | retrospective single-arm interventional | January 2004–December 2008 | USA | 1 | 46.2 ± 13.6 | 87 | 23 | 30 | 30 | 30.5 | 3 | 3 | 1 | 0 |
| 37 | Lopes, 2009* | retrospective single-arm interventional | February 2003–August 2008 | USA | 1 | 43.4 ± 7.3 | 90 | 9 | 9 | 9 | 77.8 | 4 | 0 | 0 | 0 |
| 38 | Abbas, 2018 | retrospective single-arm interventional | January 2005–December 2016 | USA, Brasil, Canada | 34 | 51 (IQR: 43–61) | 84 | 579 | 579 | 579 | 89 | 254 | 0 | 0 | 0 |
| 39 | Wisneski, 2020** | retrospective single-arm interventional | January 2012–December 2018 | USA | 1 | 53.7 (32–73) | 73 | 7 | 15 | 15 | n/a | 0 | 0 | 0 | 7 |
| 40 | Aloun, 2021** | case series | January 2015–December 2019 | Saudi Arabia | 1 | Median: 35 (26–45) | 66.6 | 1 | 3 | 3 | n/a | 1 | 0 | 0 | 0 |
| 41 | Li, 2021* | prospective cohort | January 2013–December 2018 | Germany | 1 | 48.7 ± 15.3 | 55.6 | 9 | 20 | 20 | 100 | 7 | 0 | 0 | 0 |
| 42 | Ponsky, 2017 | retrospective single-arm interventional | January 2008–October 2016 | USA | 1 | 53.9 (32–64) | 100 | 9 | 9 | 9 | 0 | 0 | 0 | 0 | 0 |
| 43 | Saleem, 2012* | retrospective single-arm interventional | February 2003–March 2010 | USA | 1 | 50.9 ± 12.6 | 80 | 15 | 15 | 15 | 33.3 | 4 | 0 | 0 | 0 |
| 44 | Habenicht, 2018 | retrospective single-arm interventional | September 2012–January 2016 | USA | 1 | 55.8 (29–67) | n/a | 16 | 16 | 16 | 100 | 15 | 0 | 0 | 0 |
| 45 | Roberts, 2008 | retrospective single-arm interventional | n/a | USA | 1 | 44.6 ± 9.5 | 100 | 5 | 5 | 5 | 20 | 1 | 0 | 0 | 0 |
| 46 | Clapp, 2021 | retrospective single-arm interventional | January 2012–December 2017 | USA | 1 | 44.8 ± 10.6 | 83.3 | 12 | 12 | 12 | n/a | n/a | n/a | n/a | n/a |
| 47 | Tzedakis, 2019* | retrospective single-arm interventional | February 2014–May 2015 | France | 1 | 42 ± 9.8 | 100 | 4 | 5 | 5 | 100 | 2 | 2 | 0 | 0 |
| 48 | Falcao, 2012* | prospective cohort | January 2003–December 2010 | Brazil | 3 | 35.3 ± 6.7 | 82.6 | 23 | 23 | 23 | 100 | 14 | 0 | 0 | 0 |
| 49 | Frederiksen, 2017 | retrospective single-arm interventional | January 2010–January 2016 | Denmark | 1 | Median: 46 (25–65) | 86 | 29 | 29 | 31 | 100 | 29 | 0 | 0 | 0 |
| 50 | Lim, 2017 | retrospective single-arm interventional | March 2011–February 2016 | USA | 1 | 50.5 (27–72) | 100 | 35 | 35 | 35 | 0 | 0 | 0 | 0 | 0 |
| 51 | Telfah, 2020 | retrospective single-arm interventional | January 2010–December 2019 | UK | 1 | 64 (34–73) | 75 | 12 | 12 | 12 | 100 | 0 | 4 | 0 | 0 |
| 52 | Koggel, 20212 | retrospective single-arm interventional | January 2009–August 2019 | The Netherlands | 1 | Median: 53.5 (27–72) | 81.4 | 86 | 86 | 100 | 93 | 61 | 25 | 0 | 0 |
| 53 | Liljegard, 2022 | case–control | January 2007–December 2020 | Sweden | 1 | Median: 58 (28–73) | 93 | 14 | 28 | 42 | 100 | 14 | 0 | 0 | 0 |
| 54 | Grimes, 2015*** | retrospective single-arm interventional | October 2004–January 2014 | USA | 1 | 48.5 (23–69) | 94 | 41 | 41 | 85 | 12 | 5 | 0 | 0 | 0 |
| 55 | Bowman, 2016 | retrospective single-arm interventional | n/a | USA | 2 | 48.8 ± 13.7 | 27 | 11 | 15 | 16 | 54.5 | 4 | 0 | 1 | 0 |
| 56 | Mohammad, 2019 | retrospective single-arm interventional | April 2008–April 2016 | USA | 1 | 54 ± 13 | 81.25 | 32 | 32 | 32 | 93 | 16 | 2 | 0 | 0 |
| 57 | AlMasri, 2022 | retrospective single-arm interventional | October 2007–July 2019 | USA | 1 | 60 (IQR: 50–67) | 80.9 | 131 | 131 | 131 | 82.5 | 102 | 0 | 3 | 0 |
| 58 | Ceppa, 2007 | retrospective single-arm interventional | July 1999–January 2005 | USA | 1 | 44 (32–56) | 80 | 5 | 10 | 10 | n/a | 4 | 0 | 0 | 0 |
| 59 | Bertin, 2011 | retrospective single-arm interventional | July 2004–January 2009 | USA | 1 | n/a | n/a | 22 | 22 | 22 | 0 | 0 | 0 | 0 | 0 |
| 60 | Levy, 2022 | retrospective single-arm interventional | April 2010—July 2020 | USA | 3 | 58.4 ± 12.2 | 84.9 | 50 | 50 | 50 | n/a | n/a | n/a | n/a | n/a |
| Comparative studies | |||||||||||||||
| 61 | Kröll, 2020 | retrospective cohort | January 2013–March 2019 | Switzerland | 1 | LA-ERCP: 45.5 (72–28); EDGE: 50.5 (49–52) | EDGE: 100; LA-ERCP: 78 | EDGE: 2; LA-ERCP: 14 | 19 | 19 | 82 | EDGE: 2, LA-ERCP: 11 | 0 | 0 | 0 |
| 62 | Kedia, 2019 | retrospective cohort | May 2005—June 2017 | USA | 3 | EDGE: 56 (35–82); LA-ERCP: 55 (33–80) | EDGE: 86; LA-ERCP: 86 | EDGE: 29; LA-ERCP: 43 | 72 | 72 | 93 | 54 (EDGE: 21, LA-ERCP: 33) | 0 | 0 | 0 |
| 63 | Kochhar, 2020 | retrospective cohort | January 2015—July 2019 | USA | 1 | EDGE: 60.77 ± 11.4; LA-ERCP: 60.78 ± 12.67; EA-ERCP: 68.58 ± 15.09 | EDGE: 77; LA-ERCP: 67; EDGE: 67 | EDGE: 26; LA-ERCP: 18; EA-ERCP: 12 | 56 | 56 | n/a | EDGE: 10, LA-ERCP: 14, EA-ERCP: 5 | EDGE: 1, LA-ERCP: 1, EA-ERCP: 2 | 0 | 0 |
| 64 | Tonnesen, 2020 | retrospective cohort | May 2013–December 2017 | Norway | 2 | EA-ERCP: 51 (37–72); LA-ERCP: 48.8 (25–77) | EA-ERCP: 84; LA-ERCP: 73 | EA-ERCP: 31; LA-ERCP: 37 | 68 | 79 | 28.5 | EA-ERCP: 17, LA-ERCP: 31 | 0 | 0 | 0 |
| 65 | Schreiner, 2012**** | retrospective cohort | September 2007—May 2011 | USA | 1 | EA-ERCP: 53; LA-ERCP: 52 | EA-ERCP: 97; LA-ERCP: 79 | EA-ERCP: 32; LA-ERCP: 24 | 56 | 56 | 88 | EA-ERCP: 10, LA-ERCP: 3 | 0 | EA-ERCP: 2, LA-ERCP: 1 | 0 |
| 66 | Bukhari, 2018 | retrospective cohort | January 2014–December 2016 | USA, Denmark | 5 | EA-ERCP: 61.8 ± 11.5; EDGE 52.5 ± 13.4 | EA-ERCP: 60; EDGE: 90 | EDGE: 30; EA-ERCP: 30 | 60 | 60 | 97 | EA-ERCP: 23, EDGE: 20 | 0 | 0 | EA-ERCP: 5, EDGE: 2 |
| 67 | Wang, 2020 | retrospective cohort | January 2009–December 2019 | USA | 2 | EDGE: 59.3 ± 6.5; LA-ERCP: 50.6 ± 15.9; EA-ERCP: 55.3 ± 14.3 | EDGE: 89; LA-ERCP: 90; EA-ERCP: 67 | EDGE: 18; LA-ERCP: 42; EA-ERCP: 70 | 130 | 130 | 87,5 | EA-ERCP: 28, LA-ERCP: 20, EDGE: 11 | EA-ERCP: 9, LA-ERCP: 5, EDGE: 0 | EA-ERCP: 7, LA-ERCP: 0, EDGE: 4 | 0 |
*Demographic data were calculated based on individual datasets available in the study
**Demographic data or results refer to the entire study population, including patients, or endoscopic procedure other than ERCP, or patients with other types of surgically altered anatomy
***Results for ERCP procedures were not directly reported in the study but were deduced by the authors from information from the text
****Indications were calculated for the number of procedures
1Mean ± SD or (range) or median (IQR) unless stated otherwise
2Results were calculated for the number of procedures
CDL, choledocholithiasis; SOD, sphincter of Oddi dysfunction; ERCP, endoscopic retrograde cholangiopancreatography; EDGE, endoscopic ultrasound-directed transgastric ERCP; LA-ERCP, laparoscopy-assisted ERCP; EA-ERCP, enteroscopy-assisted ERCP; SD, standard deviation; IQR, interquartile region
Meta-analysis
Primary Outcomes
Technical success
The technical success rate was reported in 12 studies for EDGE [13–24], 35 for LA-ERCP [14, 21, 24–56], and 8 for EA-ERCP [16, 52, 57–62]. The outcome was achieved in 96% (CI 92–98%; n = 212/214), 93% (CI 91–95%; n = 718/750), and 77% (CI 69–83%; n = 137/176) of patients undergoing EDGE, LA-ERCP, and EA-ERCP, respectively, whereas the heterogeneity measure of total I2 was not substantial (I2 = 0%, CI 0–33%). Subgroup differences were significant between the EA-ERCP and EDGE, and EA-ERCP and LA-ERCP groups (p < 0.05) (Fig. 2, S1A).
Fig. 2.
Technical success rates of EDGE, LA-ERCP and EA-ERCP. Summary forest plot of pooled proportions of technical success for each procedure and significant subgroup differences. ERCP, endoscopic retrograde cholangiopancreatography; EDGE, endoscopic ultrasound-directed transgastric ERCP; LA-ERCP, laparoscopy-assisted ERCP; EA-ERCP, enteroscopy-assisted ERCP; CI, confidence interval; I2, total heterogeneity measure
Clinical success
The clinical success was assessed in 11 studies for EDGE [13–21, 23, 24], 37 for LA-ERCP [14, 21, 24–56, 63, 64], and 10 for EA-ERCP [14, 16, 52, 57–62, 64]. ERCP was successfully performed in 93% (CI 87–97%; n = 182/188), 92% (CI 90–94%; n = 1279/1366), and 64% (CI 56–71%; n = 181/277) of the patients in the EDGE, LA-ERCP, and EA-ERCP groups, respectively. The total I2 value was not substantial (I2 = 14%, CI, 0–60%). Subgroup differences between EDGE or LA-ERCP and EA-ERCP were significant (p < 0.05) (Fig. 3, S2A).
Fig. 3.
Clinical success rates of the investigated ERCP techniques. Summary forest plot of the pooled proportions of clinical success for each procedure and significant subgroup differences. ERCP, endoscopic retrograde cholangiopancreatography; EDGE, endoscopic ultrasound-directed transgastric ERCP; LA-ERCP, laparoscopy-assisted ERCP; EA-ERCP, enteroscopy-assisted ERCP; CI, confidence interval; I2, heterogeneity measure
Secondary Outcomes
Overall early adverse event rate
A total of 11 studies reported adverse events for EDGE [13–19, 21–24], 37 for LA-ERCP [14, 21, 22, 24–40, 42–55, 63–65], and 9 for EA-ERCP [14, 16, 22, 52, 58, 59, 61, 62, 64]. Complications were diagnosed in 20% (CI 12–31%; n = 33/181), 19% (CI 14–24%; n = 280/1392), and 13% (CI 8–22%; n = 31/247) of patients in the EDGE, LA-ERCP, and EA-ERCP groups, respectively. The total I2 value was substantial (I2 = 59%, CI 0–86%). Subgroup differences were not significant (p = 0.49) (Fig. 4, S3A).
Fig. 4.
Forest plot showing the overall complication rates of each technique. Summary forest plot of pooled proportions of overall complications for each procedure and subgroup differences that were not significant. ERCP, endoscopic retrograde cholangiopancreatography; EDGE, endoscopic ultrasound-directed transgastric ERCP; LA-ERCP, laparoscopy-assisted ERCP; EA-ERCP, enteroscopy-assisted ERCP; CI, confidence interval; I2, heterogeneity measure
Procedure-related and pooled perforations
Procedure-related perforations were reported in 11 studies on EDGE [13–19, 21–24], 36 on LA-ERCP [14, 21, 22, 24–34, 36–40, 42–56, 63, 65], and 8 on EA-ERCP [16, 22, 52, 58–62]. Perforation was diagnosed in 6% (CI 3–11%; n = 5/181), 3% (CI 2–4%; n = 17/1324), and 3% (CI 1–7%; n = 1/185) of patients undergoing EDGE, LA-ERCP, and EA-ERCP, respectively, due to a specific intervention. The total I2 value was non-negligible (I2 = 3%, CI 0–35%). The subgroup differences were not statistically significant (p = 0.29). Pooled perforation rates (including ERCP-related ones) were different only in the LA-ERCP group, as was reported by 35 studies (Li et al. not included) (3%, CI 2–4%; n = 21/1315); subgroup differences remained not significant (p = 0.33) (Fig. S4A).
Overall bleeding rate
Data on the number of patients with intervention-related bleeding were provided in 11 studies for EDGE [13–19, 21–24], 33 for LA-ERCP [21, 22, 24, 26–34, 36–40, 42–56, 63], and 7 for EA-ERCP [16, 22, 52, 58, 59, 61, 62]. The bleeding rates were 5% (CI 2–9%; n = 5/181), 4% (CI 3–6%; n = 20/1101), and 8% (CI 3–17%; n = 4/146) in the EDGE, LA-ERCP, and EA-ERCP groups, respectively, whereas the total I2 was not consistent (I2 = 7%, CI 0–45%). The subgroup differences were not statistically significant (p = 0.37) (Fig. S5A).
Post-ERCP pancreatitis
Several post-ERCP pancreatitis (PEP) cases were reported in 11 studies for EDGE [13–19, 21–24], 36 for LA-ERCP [14, 21, 22, 24–34, 36–40, 42–56, 63, 65], and 8 for EA-ERCP [14, 16, 22, 52, 58, 59, 61, 62]. PEP was observed in 4% (CI 2–9%; n = 2/181), 7% (CI 6–8%; n = 74/1324), and 7% (CI 4–12%; n = 12/216) of patients undergoing EDGE, LA-ERCP, and EA-ERCP, respectively. The total I2 value was negligible (I2 = 0%; CI 0–30). Subgroup differences did not reach statistical significance (p = 0.50) (Fig. S6A).
Acute cholangitis and procedure-related mortality could not be calculated because of the scarcity of data.
Additional results and those based on comparative studies are described in the Supplementum (Document 3 and Figs. S1B–6B and S7–10).
Risk of bias Assessment
The results of the risk of bias assessment are presented in Tables S3 and S4. Comments on RoB can be found in Supplementum (Sec. 3.4.).
Publication bias
Funnel plots of publication bias and detailed comments can be found in the Supplementum (Sec. 3.3., Figs. S1A–6A and S7–10A).
Sensitivity Analysis
The results are shown in Supplementum Figures S11–S18 and remarks can be found in Supplementum, Sec. 3.5.
Certainty of Evidence
The level of evidence was very low for each outcome with indirect comparisons and low for direct comparisons. A summary of these findings is presented in Supplementum, Tables S5–S10 and Sec. 3.6.
Systematic Review
Acute Cholangitis
Acute cholangitis was reported in 11 studies for EDGE [13–19, 23, 24, 66, 67], 35 studies for LA-ERCP [14, 21, 22, 24–34, 36–40, 42–56, 63], and 9 studies for EA-ERCP [16, 22, 52, 58–62, 68]. The rates were 1/373, 8/1272, and 1/255, respectively.
Mortality
Procedure-related mortality in the study population was 0/373 for 13 studies [13–19, 21–24, 66, 67] and 1/1435 for 39 studies [14, 21, 22, 24–56, 63–65], and 0/286 for 11 studies [14, 16, 22, 52, 58–62, 64, 68] for EDGE, LA-ERCP, and EA-ERCP, respectively.
Procedure Duration for EDGE
Some studies presented mean lengths of 79 (standard deviation ( ±) 31), 49.8 (± 26.5), and 92 (± 47) minutes for EDGE in 30, 26, and 178 patients [16, 22, 66]. Another study included 29 patients and reported a mean length of 73 min (range 24–230) [21]. Kröll et al. (2020) reported a median procedure length of 101 min (interquartile range (IQR) 56–47 min), including only two patients [24].
Length of Hospital Stay for EDGE
The mean length of hospital stay was 0.8 days (range 0–5) for 29 patients, 1.61 (± 1.7) days for 26 patients, and 2.9 (± 4.6) days for 75 procedures [21, 22, 69]. Others reported a median hospital stay of 1 (IQR 1–3) day for 30 patients and 0 (IQR 0–2) day for 178 procedures [16, 66].
Length of Hospital Stay for EA-ERCP
The mean length of hospital stay was 3.26. (± 4.36) days for 12 interventions and 1.28 (range 1–8) for 19 procedures [22, 52]. Others reported a median hospital stay of 10.5 (IQR, 1.5–13) days for 30 patients [16].
Predictive Factors for Fistula Formation and the Risk of Weight Gain
In a retrospective review, 9 of 22 patients developed persistent fistulas 8 weeks after removal of 20-mm LAMS, and the LAMS dwell time was significantly longer in those with persistent fistulas than in those with spontaneous fistula closure (77 vs. 35 days, p = 0.03). It also reported unintentional weight gain for more patients with persistent fistula compared to patients with spontaneous fistula closure. Weight increase was of at least 2.5% of their preinterventional weight, but the difference between the groups was not significant though (p = 0.22) [70]. A case–control study (25 patients with persistent fistula vs. 50 patients with spontaneous fistula closure) found that LAMS dwell time ≥ 40 days was a significant predictor of persistent fistula formation (79.2%, n = 19/24 vs. 45.8%, n = 22/48; OR 4.5 (CI 1.5–14.0)). Patients who developed a fistula gained more than 5% body weight (n = 8/24, 33.3%) than those without a fistula (n = 4/39, 10.3%), which was a significant increase (OR 4.4, CI 1.2–16.7; p = 0.03) [69]. In another study, 5 of 13 patients developed a fistula following endoscopic closure [67]. James et al. found 1 patient developing persistent fistula from 11 followed-up individuals (7 were lost for follow-up). This patient gained 5.6 kg in a 7-month period, and after fistula closure, it lost weight again. The mean weight gain in all patients was 1.7 kg (SD ± 8.6 kg) after a mean 281 (± 177) days of follow-up [13].
Qualitative Assessment of Studies Not Suitable for Statistical Synthesis
A systematic review of these studies can be found in the Supplementary information.
Discussion
The selection of an appropriate ERCP procedure for RYGB patients should consider the success and complication rates of potential ERCP techniques. This study aimed to analyze these modalities comprehensively.
Technical and Clinical Success
Results of our direct and indirect comparisons suggested that the technical and clinical success rates may be significantly higher for EDGE and LA-ERCP than for EA-ERCP. The only exception was a direct comparison of LA-ERCP and EA-ERCP regarding overall adverse event rates, but the number of comparisons was minimal.
Previous reviews reported similar technical success rates for EA-ERCP, as observed in our work [7] [11], while Gkolfakis et al. (2023) reported higher technical success for EA-ERCP [10]. Studies reporting EA-ERCP excluded push enteroscopy cases; however, this did not hinder comparability with our work, as only three push enteroscopy cases were included in our study, all successful [57]. Table 2 details the results of previous meta-analyses.
Table 2.
Summary table of previous reviews
| Publication data | Number of studies | Number of procedures/patients | Technical success | Clinical success | Overall adverse events | Perforations | Bleeding | Post-ERCP pancreatitis | Procedure duration | Indications | Other information |
|---|---|---|---|---|---|---|---|---|---|---|---|
| EA-ERCP | |||||||||||
| Ayoub 2020 | 12 | 459 procedures | 80.0% (CI 71.3–87.4) | 73.2% (CI 62.5–82.6) | 6.5% (CI 3.9–9.6) | 1% (n = 6/459) | n/a | 5% (n = 23/459) 10 studies | 100.5 min (SD ± 19.2) | CDL 74% (n = 280/380) | n/a |
| Dhindsa 2020 | 5 | 205 patients | 71.4% (CI 51–85.7) | 58.7% (CI 27.6–84.1) | 8.4% (CI 5–13.6) | 1.8% (CI 0.7–4.7) | 1.5% (CI 0.4–5) | 6.3% (CI 3.7–10.4) | n/a | n/a | TS: cannulation, CS: clinical improvement and resolution |
| Klair 2020 | 10 | 398 procedures | 75.3% (CI 64.5–83.6) | 64.8% (CI 53.1–74.9) | 9.0% (CI 5.4–14.5) | n/a | n/a | n/a | n/a | n/a | CS: cannulation |
| Connel 2022 | 4 | 47 patients | n/a | 61.5% (CI 44.3–76.3) | 5.9% (n = 2) | n/a | n/a | 5.9% (n = 2) | 94.1 min | n/a | n/a |
| Gkolfakis 2023 | 55 | 6853 procedures | 87.3 (CI 85.3–89.4) | 69.1 (CI 65.3–72.9) | 5.7 (CI 4.50–6.80) | n/a | n/a | n/a | n/a | CDL 1/3, malignancy 11% (n = 753) | PEP percentage ranged between 1.7% and 4.1% and cholangitis was diagnosed in 0.2–1.6% of all cases. Other types of surgically altered anatomy were also included |
| LA-ERCP | |||||||||||
| Ayoub 2020 | 14 | 886 procedures | 98.5% (CI 97.6–99.2) | 97.9% (CI 96.7–98.7) | 19.0% (CI 12.6–26.4) | 1% (n = 10/847) | n/a | 6% (n = 53/847) | 158.4 min (SD ± 20) | 48% (408/847) | n/a |
| Dhindsa 2020 | 18 | 939 patients | 95.3% (CI 91.3–97.5) | 92.9% (CI 83.9–97.1) | 17.4% (CI 14–21.5) | 2.2% (CI 1.3–3.7) | 3.7% (CI 2.6–5.4) | 6.8% (CI 5.3–8.8) | n/a | n/a | n/a |
| Connel 2022 | 33 | 577 patients | n/a | 100% (CI 99–100) | 12.8% (n = 44) | n/a | 2.6% (n = 9) | 4.9% (n = 17) | 133.1 min | n/a | n/a |
| Gkolfakis 2023 | 18 | 767 procedures | 99.1% (CI 98.6–99.7) | 98.5% (CI 97.8–99.2) | 15.1% (CI 9.40–20.8) | n/a | n/a | n/a | n/a | n/a | PEP percentage ranged between 1.7% and 4.1% and cholangitis was diagnosed in 0.2–1.6% of all cases |
| EDGE | |||||||||||
| Dhindsa 2020 | 4 | 124 patients | 95.5% (CI 84.2–98.8) | 95.9% (CI 81.2–99.2) | 21.9% (CI 14.6–31.4) | 2.2% (CI 0.6–7.4) | 6.6% (CI 3.3–13) | 2.2% (CI 0.6–7.4) | n/a | n/a | n/a |
| Connel 2022 | 8 | 103 patients | n/a | 97% (CI 90–100) | 24.3% (n = 17) | n/a | 1.5% (n = 1) | 11.4% (n = 8) | 67.4 min | n/a | Stent dislodgement 7.1% (5), malposition 4.3% (3) |
| Gkolfakis 2023 | 9 | 253 procedures | 97.9% (CI 96.4–99.4) | 97.9% (CI 96.3–99.4) | 13.1% (CI 7.50–18.8) | n/a | n/a | n/a | n/a | n/a | PEP percentage ranged between 1.7% and 4.1% and cholangitis was diagnosed in 0.2–1.6% of all cases |
| Su 2023 | 14 | 585 procedures | 98% (CI 96–99), 13 studies | 94 (CI 90–99), 7 studies | 12% (CI 0.09–0.159), 14 studies | n/a | n/a | n/a | n/a | n/a | LAMS dislodgement 4%(95% CI 3–6), 14 studies). The overall AE rate was lower in the 20-mm than in the 15-mm LAMS group (OR = 5.79; 95% CI 2.35 to 14.29, 4 studies). No diff in access routes (6 studies) or number of stages (3 studies) |
CDL, choledocholithiasis; PEP, post-ERCP pancreatitis; LAMS, lumen-apposing metall stent; CI, confidence intervall; SD, standard deviation; TS, technical success; CS, clinical success; ERCP, endoscopic retrograde cholangiopancreatography; EDGE, endoscopic ultrasound-directed transgastric ERCP; LA-ERCP, laparoscopy-assisted ERCP; EA-ERCP, enteroscopy-assisted ERCP
Clinical success rates for EA-ERCP from previous studies were similar to ours, with only slight differences. EA-ERCP was associated with lower clinical success rates in one case [8] and similar clinical success rates in two cases compared to our study [7, 10]. Klair et al. (2020) defined clinical success as cannulation success, and their results were consistent with ours [11]. Dhindsa et al. (2020) described technical success as cannulation and clinical success as clinical improvement following therapeutic intervention [9]. These outcomes corresponded best to clinical success, as defined in our review, and their results were similar.
Reviews on the technical and clinical success of EDGE and LA-ERCP found results very similar to our meta-analysis [7–10, 12], with the addition of Dhindsa (2020) and Klair (2020) et al. defined technical and clinical success as described above. These findings can be explained by the technical differences in these ERCP methods compared to EA-ERCP, which allows for shorter and faster access routes to the papilla of Vater.
Complications
Pooled and individual complications, namely procedure-related perforations, bleeding, and PEP, may not occur significantly more frequently with EDGE or LA-ERCP than with EA-ERCP. These results contradict previous studies that found that EA-ERCP was associated with significantly lower adverse event rates than the other two techniques [7–12]. Some authors reported lower pooled complication rates for EDGE and LA-ERCP [8, 10, 12], while all previous reviews reported lower adverse event rates in the EA-ERCP group than in our review [7–12]. The methodological differences between these reviews and ours can explain this. Connel et al. (2022) investigated balloon EA-ERCP procedures only, and their primary outcome was that they focused exclusively on stone extraction. Gkolfakis et al. (2023) also reported higher technical success rates with EA-ERCP. However, most of the studies on EA-ERCP were performed on patients with other types of surgically altered anatomy. Differences in the populations may have contributed to these different results. Moreover, the results of some studies were calculated on a per-procedure basis [7, 10–12], and others may have included overlapping populations [10, 12]. In addition, the rates of individual adverse events were often not directly comparable to our results because of the limited data availability (Table 2). Furthermore, these studies defined overall complications as the sum of PEP, perforations, cholangitis, and bleeding [10, 11]. In some cases, complications were not clearly defined in advance [7, 8, 12]. Our study included a more comprehensive range of adverse events for the overall complications, notably for EDGE and LA-ERCP. Despite this, our analysis had no significant differences in the complication rates.
Limitations
Most of the included studies were retrospective in design, with several intrinsic drawbacks, such as the risk of selection and reporting bias. We tried to address them in multiple ways. Reporting bias was mitigated by enrolling only studies with consecutively collected data. Bias in patient selection was addressed by standardizing the population as much as possible. Namely, we implemented rigorous selection criteria for patients regarding anatomy (RYGB only) and the type of performed procedures (ERCP only). Further, the baseline characteristics of the included studies (Table 1) can reveal similarities regarding age, sex, and indication in each ERCP group. Multivariate analysis was not possible due to scarce data, but a qualitative analysis could also lower the risk of systematic error. According to this, patients receiving LA-ERCP tended to be younger than those in the other two groups. Beyond that, most of the indications were choledocholithiasis, but even other indications were mainly biliary (papillary diseases (mostly SOD) or biliary stenosis). Besides, most of the study populations comprised mainly female patients. In summary, the only difference in baseline characteristics may be observed in age distribution. It could be explained by the preference for less invasive procedures in the older population. In our opinion, it should not be considered as a determining factor behind procedural outcomes because none of the procedures had to be stopped or altered due to comorbidities or physical conditions of the patients. Besides, although patients undergoing EDGE or EA-ERCP tended to be older, they were still not or just slightly above the age of 65 years.
Unfortunately, bias in procedure selection of observational studies arises from unmeasured confounders regarding the indication and is induced by the physicians. Because of this, it cannot be adjusted for in statistical analysis.
Another limitation of our work was that we included studies with smaller sample sizes if they included consecutive patients to increase the amount of data. This, however, widened the confidence interval of our calculations. In addition, variations in expertise in different techniques may have affected the results. Furthermore, the QoE needed to be higher for the assessed outcomes due to deficiencies in the retrospective studies included.
Strengths
To our knowledge, this study included the largest number of studies reporting ERCP in the RYGB population. We applied a rigorous methodology for study selection and statistical synthesis. The outcomes were based on predefined criteria in our study. When possible, we performed direct comparisons associated with higher statistical values. In all but one case, the results of the direct and indirect comparisons were similar, confirming our study’s conclusion. Contrary to some previous reviews [7, 10–12], to ensure statistical independence of the data, we calculated our results based on the number of patients and not on the number of procedures. Studies with overlapping populations were excluded from the meta-analysis.
Implications for the Future
We propose some factors to be considered to help in daily clinical practice. First, an alimentary limb longer than 150 cm makes reaching the papilla very difficult with EA-ERCP. Further, LA-ERCP should be prioritized in patients with gallbladder in situ with choledocholithiasis as an indication and may be the most reasonable choice in cases of unclear chronic abdominal pain, as well [24, 52]. Beyond that, LA-ERCP or EA-ERCP may be safer choices in urgent settings, as single-session EDGE has often been associated with complications with LAMS. However, using a suturing device could make single-session EDGE safer [20]. Furthermore, multiple interventions are possible with all three interventions. In the case of EDGE, the risk of persistent fistulae formation and the associated weight gain increases with longer LAMS dwell time. If a fistula is present, argon plasma coagulation and over-the-scope clip placement or revisional surgery with gastro-gastric fistula takedown may be required for fistula closure [13, 70, 71]. An LA-ERCP procedure can also be finished by placing a G-tube in the gastrostomy, allowing for repeated ERCP [35–37, 63]. However, this carries an increased risk of gastrostomy site infections. Repeating EA-ERCP multiple times in a patient may be exhaustive and the least effective method and thus may be considered only if the other two techniques are not available locally or contraindicated (e.g., due to adhesions and inability to identify the remnant stomach with EUS).
Laparoscopic common bile duct exploration (LCBDE) for stone removal is an effective option for RYGB patients [72]. The bleeding, acute cholangitis, and pancreatitis rates of LCBDE seem to be lower than those of ERCP. However, bile leaks (2.3–16.7%) and retained stones (0–5%) are not uncommon and tend to be associated with different approaches of LCBDE (i.e., transcystic or transcholedochal). Furthermore, despite LCBDE being minimally invasive, it is still an operation with a low but still not negligible long-term risk of bile duct strictures (ca. 0.8%) and adhesions, especially if choledochotomy was performed. These facts imply that the decision between LCBDE and preoperative ERCP requires further clarification.
Percutaneous transhepatic cholangiography (PTC) is also safe in RYGB patients, but we recommend it as a salvage technique after a previously failed ERCP due to the limited intervention possibilities of the procedure.
Our results suggest that more centers should consider performing EDGE for patients with RYGB. Furthermore, managing these patients may require more specialized centers and multidisciplinary decision-making.
This study highlights the need for more prospective, controlled, multicenter trials with larger sample sizes to improve the QoE. We did a rough estimation of the sample size for an RCT where the main outcome is the complication proportion. We calculated for 10% difference with 5% first type error and 90% power, assuming a 10% loss of follow-up; the number of patients needed to be involved in such an RCT would be about 1135, which is probably far too many ever to be done. Because of that, prospective observational trials based on standardized protocols were more feasible. The optimal target population for such a trial could be bariatric RYGB patients with their gallbladder already removed and an alimentary limb shorter than 150 cm in a non-urgent setting. This concept would minimize the risk of deviation from the intended procedure.
Our results also stress the importance of the statistical independence of the data. This highlights the importance of accurate patient-based reporting of results in future studies.
Conclusion
Our findings suggest that EDGE and LA-ERCP may be better than EA-ERCP in reaching the papilla and performing successful ERCP, while they may be as safe as EA-ERCP in patients with an RYGB anatomy.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary file 1 Figure S1A. Technical success based on indirect comparison. a: Forest plot depicting proportions of technical success for each procedure and showing significant subgroup differences. b: Funnel plot showing potential publication bias. ERCP endoscopic retrograde cholangiopancreatography, EDGE endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 188 KB)
Supplementary file 2 Figure S1B. Technical success based on direct comparison. Forest plot comparing technical success between EDGE and LA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 65 KB)
Supplementary file 3 Figure S2A. Clinical success based on indirect comparison a: Forest plot depicting clinical success rates for each procedure and showing significant subgroup differences. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 189 KB)
Supplementary file 4 Figure S2B. Clinical success based on direct comparisons a: Forest plot comparing clinical success between EDGE and LA-ERCP. b: Forest plot comparing clinical success between LA-ERCP and EA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 116 KB)
Supplementary file 5 Figure S3A. Overall complication rates based on indirect comparison. a: The forest plot depicts overall complication rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 159 KB)
Supplementary file 6 Figure S3B. Overall complication rates based on direct comparison. a: Forest plot comparing overall complication rates between EDGE and LA-ERCP. b: Forest plot comparing overall complication rates between EDGE and EA-ERCP. c: Forest plot comparing overall complication rates between LA-ERCP and EA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 203 KB)
Supplementary file 7 Figure S4A. Procedure-related perforation rates with indirect comparison. a: The forest plot depicts procedure-related perforation rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 189 KB)
Supplementary file 8 Figure S4B. Procedure-related perforation rates with direct comparison. Forest plot comparing procedure-related perforation rates between EDGE and LA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 67 KB)
Supplementary file 9 Figure S5A. Proportions of overall bleeding rates with indirect comparison.a: The forest plot depicts procedure-related perforation rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure (JPG 172 KB)
Supplementary file 10 Figure S5B. Proportions of overall bleeding rates with direct comparison. Forest plot comparing overall bleeding rates between EDGE and LA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 60 KB)
Supplementary file 11 Figure S6A. Proportions of post-ERCP pancreatitis rates with indirect comparison. a: The forest plot depicts post-ERCP pancreatitis rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 191 KB)
Supplementary file 12 Figure S6B. Proportions of PEP rates with direct comparison. a: Forest plot comparing PEP rates between LA-ERCP and EA-ERCP .b: Forest plot comparing PEP rates between EDGE and LA-ERCP. c: Forest plot comparing PEP rates between EDGE and EA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, PEP: post-ERCP pancreatitis, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 181 KB)
Supplementary file 13 Figure S7. Persistent fistulae formation a: Forest plot about the rate of fistulae formation following EDGE. b: Funnel plot investigating publication bias (right). EDGE: endoscopic ultrasound-directed transgastric ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 144 KB)
Supplementary file 14 Figure S8. Complications with LAMS. A: Forest plot representing pooled complication rates occurring with LAMS (left). Funnel plot representing publication bias of the studies reporting on LAMS complications (right). B: Forest plot representing dislodgement rates occurring with LAMS (left). Funnel plot representing publication bias of the studies reporting on LAMS dislodgements (right). LAMS: lumen-apposing metall stent, CI: confidence interval, I2: total heterogeneity measure. (JPG 238 KB)
Supplementary file15 Figure S9. Length of hospital stay of patients undergoing LA-ERCP. a: Forest plot of studies reporting on length of hospital stay after LA-ERCP. b: Funnel plot of studies reporting on length of hospital stay after LA-ERCP. LA-ERCP: laparoscopy-assisted ERCP, CI: confidence interval, MRAW: untransformed (raw) means, I2: total heterogeneity measure. (JPG 146 KB)
Supplementary file16 Figure S10A. Procedure duration of LA-ERCP. a: Forest plot of studies reporting on the length of LA-ERCP procedure. b: Funnel plot showing potential publication bias. LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, MRAW: untransformed (raw) means, min: minutes, I2: total heterogeneity measure. (JPG 164 KB)
Supplementary file17 Figure S10B. Procedure duration of EA-ERCP. Forest plot of studies reporting on length of EA-ERCP procedure. EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, MRAW: untransformed (raw) means, min: minutes, I2: total heterogeneity measure. (JPG 70 KB)
Supplementary file18 Figure S11. Sensitivity analysis of indirect comparison of technical success rates using dfbetas The Logit scale represents how much the estimated effect size changes by leaving out the given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP. (JPG 69 KB)
Supplementary file 19 Figure S12. Leave-one-out sensitivity analysis values about clinical success using the confidence interval method. Sensitivity analysis of studies reporting clinical success using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 171 KB)
Supplementary file 20 Figure S13. Leave-one-out sensitivity analysis values about overall adverse event rates using the confidence interval method. Sensitivity analysis of studies reporting overall adverse events using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 164 KB)
Supplementary file 21 Figure S14. Leave-one-out sensitivity analysis values about procedure-related perforation rates using the confidence interval method. Sensitivity analysis of studies reporting on procedure-related perforations using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 160 KB)
Supplementary file 22 Figure S15. Leave-one-out sensitivity analysis values about overall bleeding rates using the confidence interval method Sensitivity analysis of studies reporting overall bleeding rates using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 152 KB)
Supplementary file 23 Figure S16. Sensitivity analysis of PEP rates with leave-one-out analysis using dfbetas. The Logit scale represents how much the estimated effect size changes, leaving out the given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, PEP: post-ERCP pancreatitis. (JPG 66 KB)
Supplementary file 24 Figure S17. Leave-one-out sensitivity analysis values about persistent fistulae development using forest plot. Sensitivity analysis of studies reporting persistent fistulae rates using estimates with confidence intervals if leaving out a given study. Effect size: the pooled effect size without the given study, CI: confidence interval, I2: heterogeneity value without the presented study, Std residual: the studentized residuals which show the deleted residual divided by its estimated standard deviation, Dffits: the difference in fits. It quantifies the number of standard deviations the fitted value changes without the given study, Cook’s dist.: Cook’s distance depends on both the residual and leverage of the omitted study, Covariance ratio: the covariance ratio showing the change in the determinant of the covariance matrix of the effect size, Hat value: the value of the hat matrix without the given study. (JPG 143 KB)
Supplementary file 25 Figure S18. Leave-one-out sensitivity analysis about complications occurring with LAMS using forest plot. a: Sensitivity analysis of studies reporting pooled complication rates with LAMS using estimates with confidence intervals if leaving out a given study. b: Sensitivity analysis of studies reporting LAMS dislodgement rates using estimates with confidence intervals if leaving out a given study. Effect size: the pooled effect size without the given study, CI: confidence interval, I2: heterogeneity value without the presented study, Std residual: the studentized residuals which show the deleted residual divided by its estimated standard deviation, Dffits: the difference in fits. It quantifies the number of standard deviations the fitted value changes without the given study, Cook’s dist.: Cook’s distance depends on both the residual and leverage of the omitted study, Covariance ratio: the covariance ratio showing the change in the determinant of the covariance matrix of the effect size, Hat value: the value of the hat matrix without the given study. (JPG 226 KB)
Author contributions
BG: conceptualization, investigation, project administration, validation, formal analysis, visualization, and writing—original draft; AR: conceptualization, methodology, project administration, and writing—review and editing; JH: investigation, validation, data curation, writing—review and editing; BT: conceptualization and writing—review and editing; ZS: conceptualization and writing—review and editing; DSV: formal analysis, data curation, visualization, and writing—review and editing; PJH: conceptualization and writing—review and editing; NF: conceptualization and writing—review and editing; SZÁ: conceptualization and writing—review and editing; PH: supervision, funding acquisition, and writing—review and editing; IH: conceptualization, supervision, and writing—original draft. All authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript.
Funding
Funding was provided by the ÚNKP-22–3 New National Excellence Program of the Ministry for Innovation and Technology from the source of the National Research, Development and Innovation Fund (to BT—ÚNKP-23–3-II-PTE-1996).
Data Availability
The data that support the findings of this study are available from the corresponding author, I.H., upon reasonable request.
Declarations
Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.
Informed consent
Informed consent does not apply.
Conflict of interest
The authors declare no conflict of interests.
Footnotes
Key points
1. Previous data on different ERCP techniques for patients following RYGB were controversial, as more effective methods appeared to have higher complication rates.
2. Our study found that the chances of successful ERCP were significantly higher with laparoscopy-assisted and endoscopic ultrasound-directed techniques than with enteroscopy-assisted ERCP.
3. Based on our results, the three methods did not significantly differ in complication rates.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Angrisani L, Santonicola A, Iovino P, et al. Bariatric surgery survey 2018: similarities and disparities among the 5 IFSO chapters. Obes Surg. 2021;31(5):1937–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gostout CJ, Bender CE. Cholangiopancreatography, sphincterotomy, and common duct stone removal via Roux-en-Y limb enteroscopy. Gastroenterology. 1988;95(1):156–63. [DOI] [PubMed] [Google Scholar]
- 3.Gray R, Leong S, Marcon N, et al. Endoscopic retrograde cholangiography, sphincterotomy, and gallstone extraction via gastrostomy. Gastrointest Endosc. 1992;38(6):731–2. [DOI] [PubMed] [Google Scholar]
- 4.Baron TH, Vickers SM. Surgical gastrostomy placement as access for diagnostic and therapeutic ERCP. Gastrointest Endosc. 1998;48(6):640–1. [DOI] [PubMed] [Google Scholar]
- 5.Kedia P, Sharaiha RZ, Kumta NA, et al. Internal EUS-directed transgastric ERCP (EDGE): game over. Gastroenterology. 2014;147(3):566–8. [DOI] [PubMed] [Google Scholar]
- 6.van der Merwe SW, van Wanrooij RLJ, Bronswijk M, et al. Therapeutic endoscopic ultrasound: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy. 2022;54(2):185–205. [DOI] [PubMed] [Google Scholar]
- 7.Ayoub F, Brar TS, Banerjee D, et al. Laparoscopy-assisted versus enteroscopy-assisted endoscopic retrograde cholangiopancreatography (ERCP) in Roux-en-Y gastric bypass: a meta-analysis. Endosc Int Open. 2020;8(3):E423–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Connell M, Sun WYL, Mocanu V, et al. Management of choledocholithiasis after Roux-en-Y gastric bypass: a systematic review and pooled proportion meta-analysis. Surg Endosc. 2022;36(9):6868–77. [DOI] [PubMed] [Google Scholar]
- 9.Dhindsa BS, Dhaliwal A, Mohan BP, et al. EDGE in Roux-en-Y gastric bypass: how does it compare to laparoscopy-assisted and balloon enteroscopy ERCP: a systematic review and meta-analysis. Endosc Int Open. 2020;8(2):E163–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gkolfakis P, Papaefthymiou A, Facciorusso A, et al. Comparison between enteroscopy-, laparoscopy- and endoscopic ultrasound-assisted endoscopic retrograde cholangio-pancreatography in patients with surgically altered anatomy: a systematic review and meta-analysis. Life (Basel). 2022 12(10)1646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Klair JS, Jayaraj M, Chandrasekar VT, et al. ERCP with overtube-assisted enteroscopy in patients with Roux-en-Y gastric bypass anatomy: a systematic review and meta-analysis. Endoscopy. 2020;52(10):824–32. [DOI] [PubMed] [Google Scholar]
- 12.Su T, Chen T, Wang J, et al. Endoscopic-directed trans-gastric retrograde cholangiopancreatography in patients with Roux-en-Y gastric bypasses: a meta-analysis. J Clin Gastroenterol. 2023;57(9):871–8. [DOI] [PubMed] [Google Scholar]
- 13.James TW, Baron TH. Endoscopic Ultrasound-Directed Transgastric ERCP (EDGE): a single-center US experience with follow-up data on fistula closure. Obesity Surgery. 2019;29(2):451–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wang TJ, Cortes P, Jirapinyo P, et al. A comparison of clinical outcomes and cost utility among laparoscopy, enteroscopy, and temporary gastric access-assisted ERCP in patients with Roux-en-Y gastric bypass anatomy. Surg Endosc. 2021;35(8):4469–77. [DOI] [PubMed] [Google Scholar]
- 15.Barclay RL, Jain A, Ferland ASB, et al. Living on the EDGE: Canadian experience with EUS-directed transgastric ERCP (EDGE) in patients with Roux-en-Y gastric bypass anatomy. J Canad Assoc Gastroenterol. 2022;5(3):116–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bukhari M, Kowalski T, Nieto J, et al. An international, multicenter, comparative trial of EUS-guided gastrogastrostomy-assisted ERCP versus enteroscopy-assisted ERCP in patients with Roux-en-Y gastric bypass anatomy. Gastrointestinal Endoscopy. 2018;88(3):486–94. [DOI] [PubMed] [Google Scholar]
- 17.Chhabra P, On W, Paranandi B, et al. Initial United Kingdom experience of endoscopic ultrasound-directed transgastric endoscopic retrograde cholangiopancreatography. Ann Hepatobiliary Pancreat Surg. 2022;26:318–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.de Benito Sanz M, Carbajo AY, Sánchez-Ocaña Hernández R, et al. Endoscopic ultrasound-directed transgastric ERCP in patients with Roux-en-Y gastric bypass using lumen-apposing metal stents or duodenal self-expandable metal stents A European single-center experience. Rev Esp Enferm Dig. 2020;112(3):211–5. [DOI] [PubMed] [Google Scholar]
- 19.Irani S, Yang J, Khashab MA. Mitigating lumen-apposing metal stent dislodgment and allowing safe, single-stage EUS-directed transgastric ERCP. VideoGIE: official Vid J Am Soc Gastrointes Endos. 2018;3(10):322–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Keane MG, Higa JT, La Selva D, et al. Same-session Endoscopic Ultrasound Directed Transgastric Intervention (EDGI) in patients with RYGB anatomy: a multicenter study. Gastrointestinal Endos. 2022;8:S0016-5107. [Google Scholar]
- 21.Kedia P, Tarnasky PR, Nieto J, et al. EUS-directed transgastric ERCP (EDGE) versus laparoscopy-assisted ERCP (LA-ERCP) for Roux-en-Y gastric bypass (RYGB) anatomy. J Clin Gastroenterol. 2019;53(4):304–8. [DOI] [PubMed] [Google Scholar]
- 22.Kochhar GS, Mohy-Ud-Din N, Grover A, et al. EUS-directed transgastric endoscopic retrograde cholangiopancreatography versus laparoscopic-assisted ERCP versus deep enteroscopy-assisted ERCP for patients with RYGB. Endoscopy International Open. 2020;8(7):E877–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Krafft MR, Fang W, Nasr JY. Shortened-interval dual-session EDGE reduces the risk of LAMS dislodgement while facilitating timely ERCP. Digest Dis Sci. 2021;66(8):2776–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kröll D, Müller AC, Nett PC, et al. Tailored access to the hepatobiliary system in post-bariatric patients: a tertiary care bariatric center experience. Surg Endosc. 2020;34(12):5469–76. [DOI] [PubMed] [Google Scholar]
- 25.AlMasri S, Zenati MS, Papachristou GI, et al. Laparoscopic-assisted ERCP following RYGB: a 12-year assessment of outcomes and learning curve at a high-volume pancreatobiliary center. Surg Endosc. 2022;36(1):621–30. [DOI] [PubMed] [Google Scholar]
- 26.Baimas-George M, Passeri MJ, Lyman WB, et al. A single-center experience with minimally invasive transgastric ERCP in patients with previous gastric bypass: lessons learned and technical considerations. Am Surg. 2020;86(4):300–7. [PubMed] [Google Scholar]
- 27.Bayoumi M, Hussein H, Aziz BAA, et al. Endoscopic retrograde cholangiopancreatography through laparoscopically created gastrotomy for the management of biliary complications of Roux-en-Y gastric bypass. Egypt J Surg. 2016;35(4):449–55. [Google Scholar]
- 28.Bertin PM, Singh K, Arregui ME. Laparoscopic transgastric endoscopic retrograde cholangiopancreatography (ERCP) after gastric bypass: case series and a description of technique. Surg Endosc. 2011;25(8):2592–6. [DOI] [PubMed] [Google Scholar]
- 29.Bowman E, Greenberg J, Garren M, et al. Laparoscopic-assisted ERCP and EUS in patients with prior Roux-en-Y gastric bypass surgery: a dual-center case series experience. Surg Endosc. 2016;30(10):4647–52. [DOI] [PubMed] [Google Scholar]
- 30.Brockmeyer JR, Grover BT, Kallies KJ, et al. Management of biliary symptoms after bariatric surgery. Am J Surg. 2015;210(6):1010–7. [DOI] [PubMed] [Google Scholar]
- 31.Ceppa FA, Gagné DJ, Papasavas PK, et al. Laparoscopic transgastric endoscopy after Roux-en-Y gastric bypass. Surg Obes Related Dis. 2007;3(1):21–4. [DOI] [PubMed] [Google Scholar]
- 32.Clapp B, Wicker E, Vivar A, Kara AM, Gamez J, Davis B. Long term outcomes after laparoscopic assisted trans-gastric endoscopic retrograde cholangiopancreatography. JSLS : J Soc Laparoendoscopic Surg. 2021;25(3):e2021.00048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Falcão M, Campos JM, Neto MG, et al. Transgastric endoscopic retrograde cholangiopancreatography for the management of biliary tract disease after Roux-en-Y gastric bypass treatment for obesity. Obes Surg. 2012;22(6):872–6. [DOI] [PubMed] [Google Scholar]
- 34.Frederiksen NA, Tveskov L, Helgstrand F, et al. Treatment of common bile duct stones in gastric bypass patients with laparoscopic transgastric endoscopic retrograde cholangiopancreatography. Obes Surg. 2017;27(6):1409–13. [DOI] [PubMed] [Google Scholar]
- 35.Grimes KL, Maciel VH, Mata W, et al. Complications of laparoscopic transgastric ERCP in patients with Roux-en-Y gastric bypass. Surg Endosc. 2015;29(7):1753–9. [DOI] [PubMed] [Google Scholar]
- 36.Gutierrez JM, Lederer H, Krook JC, et al. Surgical gastrostomy for pancreatobiliary and duodenal access following Roux en Y gastric bypass. J Gastrointes Surg : official J Soc Surg Alimen Tract. 2009;13(12):2170–5. [DOI] [PubMed] [Google Scholar]
- 37.Habenicht Yancey K, McCormack LK, McNatt SS, et al. Laparoscopic-assisted transgastric ERCP: a single-institution experience. J Obes. 2018;2018:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ivano FH, Ponte BJ, Dubik TC, et al. Endoscopic retrograde cholangiopancreatography (ERCP): analysis of the effectiveness and safety of the procedure in the patient with Roux-en-Y gastric bypass Arquivos brasileiros de cirurgia. ABCD: Braz Arch Digest Surg. 2019;32(2):e1432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Li R, Veltzke-Schlieker W, Adler A, Ismail M, et al. Intragastric single-port surgery: an innovative and multipurpose technique for the therapy of upper digestive tract lesions. Surg Innov. 2022;29(1):56–65. [DOI] [PubMed] [Google Scholar]
- 40.Liljegard S, Fredriksson Å, Manke T, Kylebäck A, Larsson PA, Haraldsson E. The outcome of laparoscopy-assisted transgastric rendezvous ERCP during cholecystectomy after Roux-en-Y gastric bypass compared to normal controls. Obes Surg. 2022;32(10):3398–402. [DOI] [PubMed] [Google Scholar]
- 41.Lim CH, Jahansouz C, Freeman ML, et al. Outcomes of endoscopic retrograde cholangiopancreatography (ERCP) and sphincterotomy for suspected sphincter of oddi dysfunction (SOD) post Roux-En-Y gastric bypass. Obes Surge. 2017;27(10):2656–62. [DOI] [PubMed] [Google Scholar]
- 42.Lopes TL, Clements RH, Wilcox CM. Laparoscopy-assisted ERCP: experience of a high-volume bariatric surgery center (with video). Gastrointes Endosc. 2009;70(6):1254–9. [DOI] [PubMed] [Google Scholar]
- 43.May D, Vogels E, Parker D, Petrick A, et al. Overall outcomes of laparoscopic-assisted ERCP after Roux-en-Y gastric bypass and sphincter of Oddi dysfunction subgroup analysis. Endosc Intl Open. 2019;7(10):E1276–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Mohammad B, Richard MN, Pandit A, et al. Outcomes of laparoscopic-assisted ERCP in gastric bypass patients at a community hospital center. Surg Endosc. 2020;34(12):5259–64. [DOI] [PubMed] [Google Scholar]
- 45.Paranandi B, Joshi D, Mohammadi B, et al. Laparoscopy-assisted ERCP (LA-ERCP) following bariatric gastric bypass surgery: Initial experience of a single UK centre. Front Gastroenterol. 2016;7(1):54–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Patel JA, Patel NA, Shinde T, et al. Endoscopic retrograde cholangiopancreatography after laparoscopic Roux-en-Y gastric bypass: a case series and review of the literature. Am Surg. 2008;74(8):689–93. [PubMed] [Google Scholar]
- 47.Ponsky JL, Jones N, Rodriguez JH, et al. Massive biliary dilation after Roux-en-Y gastric bypass: is it ampullary achalasia? J Am Coll Surg. 2017;224(6):1097–103. [DOI] [PubMed] [Google Scholar]
- 48.Popowicz A, Sanamrad S, Darkahi B, et al. Management of common bile duct stones encountered during cholecystectomy in patients with previous gastric bypass. Front Surg. 2021;8:789231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Richardson JF, Lee JG, Smith BR, et al. Laparoscopic transgastric endoscopy after Roux-en-Y gastric bypass: case series and review of the literature. Am Sur. 2012;78(10):1182–6. [PubMed] [Google Scholar]
- 50.Roberts KE, Panait L, Duffy AJ, et al. Laparoscopic-assisted transgastric endoscopy: current indications and future implications. Jsls. 2008;12(1):30–6. [PMC free article] [PubMed] [Google Scholar]
- 51.Saleem A, Levy MJ, Petersen BT, et al. Laparoscopic assisted ERCP in Roux-en-Y gastric bypass (RYGB) surgery patients. J Gastrointes Surg. 2012;16(1):203–8. [DOI] [PubMed] [Google Scholar]
- 52.Schreiner MA, Chang L, Gluck M, et al. Laparoscopy-assisted versus balloon enteroscopy-assisted ERCP in bariatric post-Roux-en-Y gastric bypass patients. Gastrointest Endosc. 2012;75(4):748–56. [DOI] [PubMed] [Google Scholar]
- 53.Snauwaert C, Laukens P, Dillemans B, et al. Laparoscopy-assisted transgastric endoscopic retrograde cholangiopancreatography in bariatric Roux-en-Y gastric bypass patients. Endosc Intl Open. 2015;3(5):E458–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Telfah MM, Noble H, Mahon D, et al. Laparoscopic-assisted endoscopic retrograde cholangiopancreatography (ERCP) for bile duct stones after Roux-en-Y-gastric bypass: single-centre experience. Obes Surg. 2020;30(12):4953–7. [DOI] [PubMed] [Google Scholar]
- 55.Tzedakis S, Memeo R, Nedelcu M, et al. Laparoscopy-assisted transgastric endoscopic retrograde cholangiopancreatography: preliminary experience and technique description. J Visceral Surg. 2019;156(5):381–6. [DOI] [PubMed] [Google Scholar]
- 56.Wisneski AD, Carter J, Nakakura EK, et al. Ampullary stenosis and choledocholithiasis post Roux-En-Y gastric bypass: challenges of biliary access and intervention. HPB. 2020;22(10):1496–503. [DOI] [PubMed] [Google Scholar]
- 57.Elton E, Hanson BL, Qaseem T, et al. Diagnostic and therapeutic ERCP using an enteroscope and a pediatric colonoscope in long-limb surgical bypass patients☆☆☆★. Gastrointest Endosc. 1998;47(1):62–7. [DOI] [PubMed] [Google Scholar]
- 58.Emmett DS, Mallat DB. Double-balloon ERCP in patients who have undergone Roux-en-Y surgery: a case series. Gastrointest Endosc. 2007;66(5):1038–41. [DOI] [PubMed] [Google Scholar]
- 59.Sawas T, Storm AC, Bazerbachi F, et al. An innovative technique using a percutaneously placed guidewire allows for higher success rate for ERCP compared to balloon enteroscopy assistance in Roux-en-Y gastric bypass anatomy. Surg Endos. 2020;34(2):806–13. [DOI] [PubMed] [Google Scholar]
- 60.Siddiqui AA, Chaaya A, Shelton CE, et al. Utility of the short double-balloon enteroscope to perform pancreaticobiliary interventions in patients with surgically altered anatomy in a US multicenter study. Digest Dis Sci. 2012;58(3):858–64. [DOI] [PubMed] [Google Scholar]
- 61.Trindade AJ, Mella JM, Slattery E, et al. Use of a cap in single-balloon enteroscopy-assisted endoscopic retrograde cholangiography. Endoscopy. 2015;47(5):453–6. [DOI] [PubMed] [Google Scholar]
- 62.Wagh MS, Draganov PV. Prospective evaluation of spiral overtube-assisted ERCP in patients with surgically altered anatomy. Gastrointest Endosc. 2012;76(2):439–43. [DOI] [PubMed] [Google Scholar]
- 63.Abbas AM, Strong AT, Diehl DL, et al. Multicenter evaluation of the clinical utility of laparoscopy-assisted ERCP in patients with Roux-en-Y gastric bypass. Gastrointest Endosc. 2018;87(4):1031–9. [DOI] [PubMed] [Google Scholar]
- 64.Tonnesen CJ, Young J, Glomsaker T, et al. Laparoscopy-assisted versus balloon enteroscopy-assisted ERCP after Roux-en-Y gastric bypass. Endoscopy. 2020;52(8):654–61. [DOI] [PubMed] [Google Scholar]
- 65.Levy A, Smith C, St. Hill C, et al. Complications rate of surgically assisted endoscopic retrograde cholangiopancreatography. Surg Pract Sci 2022 9:100092 [Google Scholar]
- 66.Runge TM, Chiang AL, Kowalski TE, et al. Endoscopic ultrasound-directed transgastric ERCP (EDGE): a retrospective multicenter study. Endoscopy. 2021;53(6):611–8. [DOI] [PubMed] [Google Scholar]
- 67.Tyberg A, Kedia P, Tawadros A, et al. EUS-directed transgastric endoscopic retrograde cholangiopancreatography (EDGE): the first learning curve. J Clin Gastroenterol. 2020;54(6):569–72. [DOI] [PubMed] [Google Scholar]
- 68.Ali MF, Modayil R, Gurram KC, et al. Spiral enteroscopy–assisted ERCP in bariatric-length Roux-en-Y anatomy: a large single-center series and review of the literature (with video). Gastrointest Endosc. 2018;87(5):1241–7. [DOI] [PubMed] [Google Scholar]
- 69.Ghandour B, Keane MG, Shinn B, et al. Factors predictive of persistent fistulas in EUS-directed transgastric ERCP: a multicenter matched case-control study. Gastrointest Endosc. 2022;97:260–6. [DOI] [PubMed] [Google Scholar]
- 70.Krafft MR, Lorenze A, Croglio MP, et al. “Innocent as a LAMS”: does spontaneous fistula closure (secondary intention), after EUS-directed transgastric ERCP (EDGE) via 20-mm lumen-apposing metal stent, confer an increased risk of persistent fistula and unintentional weight gain? Digest Dis Sci. 2022;67(6):2337–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 71.Ghandour B, Shinn B, Dawod QM, et al. EUS-directed transgastric interventions in Roux-en-Y gastric bypass anatomy: a multicenter experience. Gastrointest Endosc. 2022;96(4):630–8. [DOI] [PubMed] [Google Scholar]
- 72.Zerey M, Haggerty S, Richardson W, et al. Laparoscopic common bile duct exploration. Surg Endosc. 2018;32(6):2603–12. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary file 1 Figure S1A. Technical success based on indirect comparison. a: Forest plot depicting proportions of technical success for each procedure and showing significant subgroup differences. b: Funnel plot showing potential publication bias. ERCP endoscopic retrograde cholangiopancreatography, EDGE endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 188 KB)
Supplementary file 2 Figure S1B. Technical success based on direct comparison. Forest plot comparing technical success between EDGE and LA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 65 KB)
Supplementary file 3 Figure S2A. Clinical success based on indirect comparison a: Forest plot depicting clinical success rates for each procedure and showing significant subgroup differences. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 189 KB)
Supplementary file 4 Figure S2B. Clinical success based on direct comparisons a: Forest plot comparing clinical success between EDGE and LA-ERCP. b: Forest plot comparing clinical success between LA-ERCP and EA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 116 KB)
Supplementary file 5 Figure S3A. Overall complication rates based on indirect comparison. a: The forest plot depicts overall complication rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 159 KB)
Supplementary file 6 Figure S3B. Overall complication rates based on direct comparison. a: Forest plot comparing overall complication rates between EDGE and LA-ERCP. b: Forest plot comparing overall complication rates between EDGE and EA-ERCP. c: Forest plot comparing overall complication rates between LA-ERCP and EA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 203 KB)
Supplementary file 7 Figure S4A. Procedure-related perforation rates with indirect comparison. a: The forest plot depicts procedure-related perforation rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 189 KB)
Supplementary file 8 Figure S4B. Procedure-related perforation rates with direct comparison. Forest plot comparing procedure-related perforation rates between EDGE and LA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 67 KB)
Supplementary file 9 Figure S5A. Proportions of overall bleeding rates with indirect comparison.a: The forest plot depicts procedure-related perforation rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure (JPG 172 KB)
Supplementary file 10 Figure S5B. Proportions of overall bleeding rates with direct comparison. Forest plot comparing overall bleeding rates between EDGE and LA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 60 KB)
Supplementary file 11 Figure S6A. Proportions of post-ERCP pancreatitis rates with indirect comparison. a: The forest plot depicts post-ERCP pancreatitis rates for each procedure and shows that subgroup differences were not significant. b: Funnel plot exploring publication bias for each study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 191 KB)
Supplementary file 12 Figure S6B. Proportions of PEP rates with direct comparison. a: Forest plot comparing PEP rates between LA-ERCP and EA-ERCP .b: Forest plot comparing PEP rates between EDGE and LA-ERCP. c: Forest plot comparing PEP rates between EDGE and EA-ERCP. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, PEP: post-ERCP pancreatitis, OR: odds ratio, CI: confidence interval, I2: total heterogeneity measure. (JPG 181 KB)
Supplementary file 13 Figure S7. Persistent fistulae formation a: Forest plot about the rate of fistulae formation following EDGE. b: Funnel plot investigating publication bias (right). EDGE: endoscopic ultrasound-directed transgastric ERCP, CI: confidence interval, I2: total heterogeneity measure. (JPG 144 KB)
Supplementary file 14 Figure S8. Complications with LAMS. A: Forest plot representing pooled complication rates occurring with LAMS (left). Funnel plot representing publication bias of the studies reporting on LAMS complications (right). B: Forest plot representing dislodgement rates occurring with LAMS (left). Funnel plot representing publication bias of the studies reporting on LAMS dislodgements (right). LAMS: lumen-apposing metall stent, CI: confidence interval, I2: total heterogeneity measure. (JPG 238 KB)
Supplementary file15 Figure S9. Length of hospital stay of patients undergoing LA-ERCP. a: Forest plot of studies reporting on length of hospital stay after LA-ERCP. b: Funnel plot of studies reporting on length of hospital stay after LA-ERCP. LA-ERCP: laparoscopy-assisted ERCP, CI: confidence interval, MRAW: untransformed (raw) means, I2: total heterogeneity measure. (JPG 146 KB)
Supplementary file16 Figure S10A. Procedure duration of LA-ERCP. a: Forest plot of studies reporting on the length of LA-ERCP procedure. b: Funnel plot showing potential publication bias. LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, MRAW: untransformed (raw) means, min: minutes, I2: total heterogeneity measure. (JPG 164 KB)
Supplementary file17 Figure S10B. Procedure duration of EA-ERCP. Forest plot of studies reporting on length of EA-ERCP procedure. EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval, MRAW: untransformed (raw) means, min: minutes, I2: total heterogeneity measure. (JPG 70 KB)
Supplementary file18 Figure S11. Sensitivity analysis of indirect comparison of technical success rates using dfbetas The Logit scale represents how much the estimated effect size changes by leaving out the given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP. (JPG 69 KB)
Supplementary file 19 Figure S12. Leave-one-out sensitivity analysis values about clinical success using the confidence interval method. Sensitivity analysis of studies reporting clinical success using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 171 KB)
Supplementary file 20 Figure S13. Leave-one-out sensitivity analysis values about overall adverse event rates using the confidence interval method. Sensitivity analysis of studies reporting overall adverse events using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 164 KB)
Supplementary file 21 Figure S14. Leave-one-out sensitivity analysis values about procedure-related perforation rates using the confidence interval method. Sensitivity analysis of studies reporting on procedure-related perforations using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 160 KB)
Supplementary file 22 Figure S15. Leave-one-out sensitivity analysis values about overall bleeding rates using the confidence interval method Sensitivity analysis of studies reporting overall bleeding rates using estimates with confidence intervals if leaving out a given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, CI: confidence interval. (JPG 152 KB)
Supplementary file 23 Figure S16. Sensitivity analysis of PEP rates with leave-one-out analysis using dfbetas. The Logit scale represents how much the estimated effect size changes, leaving out the given study. ERCP: endoscopic retrograde cholangiopancreatography, EDGE: endoscopic ultrasound-directed transgastric ERCP, LA-ERCP: laparoscopy-assisted ERCP, EA-ERCP: enteroscopy-assisted ERCP, PEP: post-ERCP pancreatitis. (JPG 66 KB)
Supplementary file 24 Figure S17. Leave-one-out sensitivity analysis values about persistent fistulae development using forest plot. Sensitivity analysis of studies reporting persistent fistulae rates using estimates with confidence intervals if leaving out a given study. Effect size: the pooled effect size without the given study, CI: confidence interval, I2: heterogeneity value without the presented study, Std residual: the studentized residuals which show the deleted residual divided by its estimated standard deviation, Dffits: the difference in fits. It quantifies the number of standard deviations the fitted value changes without the given study, Cook’s dist.: Cook’s distance depends on both the residual and leverage of the omitted study, Covariance ratio: the covariance ratio showing the change in the determinant of the covariance matrix of the effect size, Hat value: the value of the hat matrix without the given study. (JPG 143 KB)
Supplementary file 25 Figure S18. Leave-one-out sensitivity analysis about complications occurring with LAMS using forest plot. a: Sensitivity analysis of studies reporting pooled complication rates with LAMS using estimates with confidence intervals if leaving out a given study. b: Sensitivity analysis of studies reporting LAMS dislodgement rates using estimates with confidence intervals if leaving out a given study. Effect size: the pooled effect size without the given study, CI: confidence interval, I2: heterogeneity value without the presented study, Std residual: the studentized residuals which show the deleted residual divided by its estimated standard deviation, Dffits: the difference in fits. It quantifies the number of standard deviations the fitted value changes without the given study, Cook’s dist.: Cook’s distance depends on both the residual and leverage of the omitted study, Covariance ratio: the covariance ratio showing the change in the determinant of the covariance matrix of the effect size, Hat value: the value of the hat matrix without the given study. (JPG 226 KB)
Data Availability Statement
The data that support the findings of this study are available from the corresponding author, I.H., upon reasonable request.





