4670 Endoscopic therapy in 207 bile leaks: validation of a treatment strategy

Autor: Roland Ter, Gurpal S. Sandha, Paul Kortan, Michael J. Bourke, Gregory B. Haber, Christine Ross, Ryan Ponnudurai, Wellesley Site, Donald G. Ormonde
Rok vydání: 2000
Předmět:
Zdroj: Gastrointestinal Endoscopy. 51:AB198
ISSN: 0016-5107
DOI: 10.1016/s0016-5107(00)14517-1
Popis: Background: Bile leak is the most common post-cholecystectomy complication with a variety of highly successful endoscopic therapies. In 1994, one author (MB) proposed a grading system to distinguish the severity of leak into low-grade (LG, leak identified only after intra-hepatic filling of contrast) or high-grade (HG, leak seen prior to intrahepatic filling). Subsequent therapy was based on this classification with sphincterotomy (BS) alone in LG and a stent (St) to bridge the leak in HG. Results: During a 10 yr period, 207 pts (mean age 54 yrs, 127 F) with bile leaks were referred to our unit for endoscopic management. Of these, 134 underwent laparoscopic cholecystectomy, 72 open cholecystectomy and 1 had spontaneous rupture of the bile duct. Pts presented 11.8 d (mean, range 1-50 d) after surgery. Modes of presentation were leakage of bile (drains 37%, T tube 11%, incisions 12%), pain (56%), jaundice (16%), fever (11%) and abdominal distension (7%). ERCP identified the site of leak in 204 pts: cystic duct stump in 159 (78%), duct of Luschka in 26 (13%) and other in 19 (9%). A review of the initial 85 pts classified leaks into LG and HG (see above). BS alone for LG leaks and St for HG leaks proved effective in 82/85 pts (96%). This strategy was then prospectively validated in the subsequent 122 pts. Results of the combined group (104 LG, 100 HG) are presented here. In the LG group, 75/104 pts had BS alone with improvement in 68 pts (91%). Further treatment was required in 7 pts (6 had St, 1 underwent surgery). St was the initial treatment in the remaining 29/104 pts. The reasons for this were: CBD stricture (11/29), coagulopathy preventing BS (8/29), severe sepsis (3/29), prior BS with inadequate drainage (2/29) and unclear indications (5/29). In the HG group, 97/100 pts had St. Persistent leakage necessitated re-stenting in 4/97 pts. Healing of the leak was documented on follow-up ERCP in all 97 pts. Three pts were not amenable to endoscopic therapy (2 with clips across CBD and 1 with incomplete cholecystectomy) and were referred for surgery. CBD stones were identified in 41 pts (28/104 LG, 13/100 HG) and extracted in all. The distribution of stone cases indicates no impact on severity of leak. Only 3 complications occured: 2 pts developed post-ERCP pancreatitis and 1 had duodenal perforation. There was no mortality. Conclusion: The use of this simple grading system for bile leaks and the relevent therapy has proven to be a useful tool for choice of endotherapy. The role of a stent without sphincterotomy or naso-biliary tube alone has not been evaluated in this series.
Databáze: OpenAIRE
Externí odkaz: