Carl Chartrand-Lefebvre, MD #{149} Michel-Pierre Dufresne, MD, FRCPC #{149} Michel Lafortune, MD, FRCPC Real Lapointe, MD, FRCSC #{149} Michel Dagenais, MD, FRCSC #{149} Andr#{233}Roy, MD, FRCSC latrogenic Injury to the Bile Duct: A Working Classification for Radiologists’ 523 PURPOSE: To assess the usefulness of the Bismuth classification method in the preoperative localization of iatrogenic bile duct lesions with chol- angiography and to correlate these cholangiographic findings with sur- gical findings. MATERIALS AND METHODS: The records of 33 patients who un- derwent open or laparoscopic chole- cystectomy and who sustained inju- ries to the biiary tract during the course of these procedures were re- viewed retrospectively. The accuracy of the cholangiographic localization of bile duct injury was assessed with the Bismuth classification method, which is based on the localization of the traumatic lesion according to the distance from the biliary confluence. RESULTS: An exact correspondence between cholangiographic and surgi- cal findings was found in 85% of the subjects. A minimal discrepancy was found in the remainder. There was no interobserver variation. CONCLUSION: The use of the Bis- muth classification method appears to be an accurate and practical meth- od for the grading of postoperative bile duct lesions with cholangiogra- phy. Index terms: Bile ducts, injuries, 76.458 #{149} Bile ducts, leakage, 76.458 #{149} Bile ducts, stenosis or obstruction, 76.458 #{149} Bile ducts, surgery, 76.458 Radiology 1994; 193:523-526 I From the Departments of Radiology (CCL., M.P.D., ML.) and Surgery (R.L., M.D., AR.), Ho- pital Saint-Luc, 1058, rue Saint-Denis, Montreal, Quebec, Canada H2X 3J4. Received December 21, 1993; revision requested February 14, 1994; revision received May 10; accepted May 16. Ad- dress reprint requests to M.P.D. C RSNA, 1994 C HOLECYSTECTOMY is associated with a risk of iatrogenic bile duct in- jury (1-5). To our knowledge, how- ever, the radiology literature offers no systematic, tested method of describ- ing these postcholecystectomy bile duct lesions (4,6-11). The location of a bile duct injury is important for analy- sis of treatment options and predic- tion of patient outcome (12,13). A simple anatomic method for the clas- sification of postoperative bile duct stenosis was developed by Bismuth (14,15), and its use is widely discussed in the surgery literature (14-19). The aim of this study was to assess the usefulness of this classification meth- od in the preoperative localization of iatrogenic bile duct lesions with chol- angiography. We retrospectively reviewed the records of a series of patients who underwent open or laparoscopic cho- lecystectomy and who sustained in- juries to the biliary tract during the course of these procedures. The accu- racy of the cholangiographic localiza- tion of bile duct injury was assessed by comparison of these cholangio- graphic findings with surgical find- ings. MATERIALS AND METHODS Between April 1971 and February 1993, 33 patients (11 men and 22 women; mean age, 51 years ± 15) with bile duct injury from previous cholecystectomy (27 open and six laparoscopic procedures) were evaluated and underwent surgical biliary repair (median year, 1988). Eleven patients had a stenotic anastomosis that resulted from previous surgical repair of a postop- erative benign lesion. One patient also un- derwent percutaneous biliary dilation be- cause of a stricture that recurred after surgical repair. The location of the postoperative biliary lesion was identified with percutaneous transhepatic or tube cholangiography in 28 patients and with endoscopic retro- grade cholangiography in five patients. Three of the 33 patients underwent both percutaneous transhepatic cholangiogra- phy and endoscopic retrograde cholangi- ography. The reasons for the use of en- doscopic retrograde cholangiography induded a suspected traumatic cause for a biliary obstruction that still remained to be differentiated from other causes, a pri- mary referral to a gastroenterologist, and an associated distal choledocholithiasis. Percutaneous transhepatic cholangiog- raphy was performed from a right inter- costal approach with a 22-gauge Chiba needle (Cook, Bloomington, md). External biiary drainage was started in two pa- tients before the cholangiogram was ob- tamed. Four patients required percutane- ous drainage of excess bile, which was accomplished under the guidance of ultra- sonography (US) or computed tomogra- phy (CT). Injuries were graded according to the Bismuth classification method for bile duct injury (14-16). This dassification method is based on the level of the traumatic le- sion in relation to the confluence of the hepatic ducts (Fig 1). A type 1 injury is a traumatic lesion of the main bile duct, which is located more than 2 cm distal from the hepatic bifurcation (Fig 2). A type 2 injury is situated within 2 cm distal of the bifurcation (Fig 3). For type 1 and 2 injuries, the distance from the bifurcation was measured between the lowest point of the bifurcation and the highest point of the lesion. A type 3 injury completely de- stroys the hepatic stump distal to the bi- furcation but leaves the bifurcation intact (Fig 4). A type 4 injury partially or corn- pletely obliterates the bifurcation (Fig 5). A type 5 injury involves an injury of a van- ant right hepatic branch or of the main bile duct at its junction with the variant duct (Fig 6). By using the Bismuth classification method, the preoperative cholangiogram of each patient was retrospectively re- viewed by two independent observers (C.C.L., M.P.D.). The traumatic lesion of the bile duct was graded according to the standards of one of the aforementioned types. For the three patients who had un- dergone both percutaneous transhepatic cholangiography and endoscopic retro- grade cholangiography, the cholangio- graphic type was determined from the findings of the former modality. For each patient, the lesion was also graded on the