Efficacy & safety of EUS-directed transgastric endoscopic retrograde cholangiopancreatography (EDGE) in Roux-en-Y gastric bypass anatomy: a systematic review & meta-analysis
Smit S Deliwala,Babu P Mohan,Pradeep Yarra,Shahab R Khan,Saurabh Chandan,Daryl Ramai,Lena L Kassab,Antonio Facciorusso,Manish Dhawan,Douglas G Adler,Vivek Kaul,Saurabh Chawla,Gursimran S Kochhar
DOI: https://doi.org/10.1007/s00464-023-09926-7
Abstract:Background: In patients with Roux-en-Y gastric bypass (RYGB) anatomy, laparoscopic endoscopic retrograde cholangiopancreatography (LA-ERCP) and enteroscopy-assisted ERCP (E-ERCP) have been utilized to achieve pancreaticobiliary access. Endoscopic ultrasound-directed transgastric endoscopic retrograde cholangiopancreatography (EDGE) has recently emerged as an alternate and efficient approach. As data regarding EDGE continues to evolve, concerns about safety and efficacy remain, limiting wide adoptability. We performed a systematic review and meta-analysis to assess the safety and efficacy of EDGE and compare it to the current standard of care. Methods: A comprehensive search of major databases (inception to Nov 2022) identified published studies on EDGE. A random-effects model was used to calculate the pooled rates and heterogeneity (I2). Risk ratio (RR) and standardized difference in means (SMD) were utilized for head-to-head comparison analysis between EDGE vs. LA-ERCP and EDGE vs. E-ERCP. Primary outcomes assessed pooled EDGE safety (adverse events) and efficacy (technical/clinical success). Secondary outcomes assessed efficacy and safety profiles via a comparative analysis of EDGE vs. LA-ERCP and EDGE vs. E-ERCP. Results: A total of 16 studies (470 patients) were included. EDGE pooled technical success (TS) rate was 96% (95% CI 92-97.6, I2 = 0), and clinical success was 91% (85-95, I2 = 0). Pooled rate of all adverse events with EDGE was 17% (14-24.6, I2 = 32%). On sub-group analysis, these included failure of fistula closure 17% (10-25.5, I2 = 48%), stent migration 7% (4-12, I2 = 51%), bleeding 5% (3.2-7.9, I2 = 0), post-EDGE weight gain 4% (2-9, I2 = 0), perforation 4% (2.1-5.8, I2 = 0), and post-ERCP pancreatitis 2% (1-5, I2 = 0). EDGE TS was comparable to LA-ERCP (97% vs. 98%; RR, 1.00; CI, 0.85-1.17, p = 0.95) and E-ERCP (100% vs. 66%; RR, 1.26; CI, 0.99-1.6, p = 0.06). No statistical difference was noted in adverse events between EDGE and LA-ERCP (13% vs. 17.6%; RR, 0.61; CI, 0.28-1.35, p = 0.52) and E-ERCP (9.6% vs. 16%; RR, 0.61; CI, 0.28-1.35, p = 0.22). EDGE procedure time and hospital stay were shorter than LA-ERCP and E-ERCP (p < 0.001). Conclusion: Our analysis shows that EDGE is safe and efficacious to the current standard of care. Further head-to-head comparative trials are needed to validate our findings.