Biliary leaks after laparoscopic cholecystectomy: time to stent or time to drain.

BACKGROUND: Endoscopic retrograde cholangiopan-creatography (ERCP) with placement of a biliary stent or nasobiliary (NB) drain is the procedure of choice for treatment of post-cholecystectomy bile duct leaks. The aim of this study was to compare the effect of NB drainage versus internal biliary stenting on rates of leak closure, time elapsed until drain or stent removal, length of hospital stay and number of required endoscopic procedures.
METHODS: Charts were reviewed on 20 patients who underwent laparoscopic cholecystectomy complicated by Luschka or cystic duct leak. Ten patients were treated with NB drains connected to low intermittent suction and repeat NB cholangiograms were performed until leak closure was observed. Ten patients were treated with internal biliary stents. Biliary sphincterotomies were performed for stone extraction or a presumed papillary stenosis. Large bilomas were drained percutaneously prior to stenting.
RESULTS: In all 20 patients, a cholangiogram and successful placement of a NB drain or internal stent was achieved. Four patients (20%) were found to have bile duct stones, which were extracted following a sphincterotomy. Sixteen patients required percutaneous drains to evacuate large bilomas prior to biliary instrumentation. Fifteen cystic duct leaks and 5 Luschka duct leaks were reviewed. There were no complications related to ERCP. Closure of the leak was documented within 2 to 11 days (mean 4.7+/-0.9 days) in patients receiving a NB drain. The drains were removed non-endoscopically following leak closure. The internal stent group required stenting for 14 to 53 days (mean 29.1+/-4.4 days). The stent was then removed endoscopically after documentation of leak closure. Bile leaks following laparoscopic cholecystectomy closed rapidly after NB drainage and did not require repeat endoscopy for removal of the NB drain, resulting in fewer ERCPs required for treatment of biliary leaks. Internal biliary stents were in place longer owing to the nature of this intermittent endoscopic approach but an accurate comparison of time to leak closure could not be determined. Leak closure resulted once the bile flow was re-established, regardless of the technique, but removal of the NB drains was performed earlier than removal of the biliary stents. The number of ERCPs required per patient was 1.0+/-0 in the NB group and 2.2+/-0.1 (range 2-3) in the internal stent group. The length of hospitalization was 8.7+/-3.3 days for the NB group and 7.5+/-2.3 days for the internal stent group. Biliary stent placement resulted in an insignificant decrease in hospitalization at the expense of generating twice as many endoscopic procedures.
CONCLUSIONS: Our data suggest that NB drainage may be advantageous in patients requiring a prolonged hospital admission or in patients in whom repeat endoscopy is undesirable. Internal biliary stenting appears preferable when early discharge is anticipated or when expertise in placement and management of NB drains is lacking.