Endoscopic management of simple & complicated biliary leaks

Most "simple" biliary leaks are successfully managed by transpapillary stenting [TPS] and/or endoscopic sphincterotomy [ES]. Biliary leaks with concomitant obstruction to flow by stricture or stone have been termed "complex"; and typically require adjunctive therapies. We have identified a subgroup of bile leaks distinguished by being "complicated" by a local organized fluid collection and their poor response to "simple" management. We define these complicated bile leaks in the context of all leaks and describe specific management strategies. Methods: We reviewed all bile duct leaks identified and managed endoscopically over a 40 month period. We reviewed the etiology, site of ductal injury, presence of focal fluid collection, type of endoscopic intervention and clinical outcome. Results: Among 36 patients, the etiologies were: laparoscopic cholecystectomy(12), open cholecystectomy(3). T-tube removal after orthotopic liver transplantation [OLT](12), biliary anastomosis after OLT(3), abdominal trauma(3), partial hepatectomy for malignancy(2), and other abdominal surgery(l). Endoscopic management included TPS (21), ES(1), combined TPS and ES(12), and 7Fr nasobiliary [NB] tube(2) (both for OLT T-tube tract leaks). Of those with TPS, 3 of the stents were preferentially placed to bridge the leak. Transpapillary stent diameters were 10 Fr(12) and 7 Fr(20). Follow up ERCP was performed to confirm leak healing (mean interval = 4.5 wk) except after NB tube placement. At follow up ERCP 28 patients(78%) had resolution of the leak. Eight patients had persistant leaks: i)Two post-laparoscopic cholecystectomy cystic duct leaks initially treated with 7 Fr TPS had persistant catheter drainage at I week, prompting repeat TPS with 12Fr stent fascillitating resolution at 4 weeks follow up. ii) Three patients with focal flud collections at the leak site (biloma=2, hemutoma=1) failed to respond to 7F TPS's at 2,4, and 6 wks respectively. In these three cases two large diameter (≥ 10Fr) TPS's were placed in parallel bridging the leak site, and follow up ERCP at 5 to 9 wk confirmed resolution, iii) One post-OLT T-tube tract leak was repaired surgically 3 days after ERCP.iv) Two patients had non-biliary sepsis and died of septic complications. Among patients who responded to initial therapy, local fluid collection at the leak site was only documented in 6 patients. Conclusions: (1) "Simple" bile leaks are effectivly managed by TPS, ES, TPS + ES, or NB tube in the majority of cases. (2) TPS with large diameter (≥ 10 Fr) stent alone achieves high success and avoids the risks of ES. (3) "Complicated" bile leaks may be effectivly managed by placing two stents in parallel, bridging the leak.