医学专业英语.doc

Differentiating Malignant from Benign Common Bile Duct Stricture with Multiphasic Helical CT1 Seung Hong Choi, MD, Joon Koo Han, MD, Jeong Min Lee, MD, Kyoung Ho Lee, MD, Se Hyung Kim, MD, Jae Young Lee, MD and Byung Ihn Choi, MD 1 From the Department of Radiology and the Institute of Radiation Medicine, Seoul National University College of Medicine, Clinical Research Institute, Seoul National University Hospital, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea. Received May 2, 2004; revision requested July 14; revision received August 3; accepted September 23. Supported in part by the 2001 BK21 Project for Medicine, Dentistry, and Pharmacy. Address correspondence to J.K.H. (e-mail: hanjkradcom.snu.ac.kr ). ABSTRACT TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferencesPURPOSE: To evaluate retrospectively the use of multiphasic helical computed tomography (CT) to differentiate malignant and benign common bile duct (CBD) strictures in patients with only a focal CBD stricture and to determine predictors for this differentiation. MATERIALS AND METHODS: Institutional review board approval and informed patient consent were not required. Fifty patients (35 men, 15 women; age range, 3587 years; mean age, 61.6 years) with only a focal CBD stricture comprised the sample for this study (32 malignant and 18 benign strictures). The diagnosis of all malignant and five benign CBD strictures was confirmed by reviewing patients surgical and pathology records; in 13 benign CBD strictures, the diagnosis was confirmed by means of clinical features. Multiphasic CT findings were analyzed with regard to the wall thickness, location, length, and enhancement pattern of the involved CBD, the upstream CBD diameter, and other findings. CT features to identify benign and malignant CBD strictures were compared by means of univariate analysis and multivariable stepwise logistic regression analysis. RESULTS: Malignant strictures were longer (17.9 mm 6.6 standard deviation) than benign strictures (8.9 mm 6.8) (P .0001), and upstream CBD diameters were larger in malignant cases (22.0 mm 5.4) than in benign cases (17.8 mm 4.6) (P = .033). The involved wall thickness was more than 1.5 mm in 26 malignant cases and three benign cases (P .0001). During both hepatic arterial and portal venous phases, greater enhancement than that in the normal CBD were more frequently observed in malignant cases (in 27 and 30 patients for hepatic arterial and portal venous phase scans, respectively) than in benign cases (in two and three patients, respectively) (P .0001). Results of multivariable stepwise logistic regression analysis showed that hyperenhancement of the involved CBD during the portal venous phase was the only variable that could be used to independently differentiate malignant from benign strictures. CONCLUSION: Hyperenhancement of the involved CBD during the portal venous phase is the main factor distinguishing malignant from benign CBD strictures. RSNA, 2005 INTRODUCTION TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferencesCholangiocarcinoma is an adenocarcinoma that arises from the bile duct epithelium. It is usually classified as either intrahepatic or extrahepatic. Cholangiocarcinoma typically manifests in one of three shapes: infiltrative, exophytic, or polypoid. Most extrahepatic cholangiocarcinomas are of the infiltrative type and manifest as a focal stricture of the bile duct, whereas papillary carcinoma is occasionally reported to manifest as an intraductal polypoid mass (14). Benign common bile duct (CBD) strictures, which have several possible causes (eg, cholangitis, traumatic and postsurgical sequelae, chronic pancreatitis, and papillary stenosis), may mimic infiltrative cholangiocarcinoma (5). At cholangiography, typical malignant CBD strictures usually manifest as irregular strictures with a shouldered margin, whereas benign stenoses tend to have smooth borders with tapered margins (6). In addition, computed tomographic (CT) findings, including thickening and enhancement of involved CBD, extrabiliary manifestations such as lymph node metastasis, and distant metastasis, enable malignant strictures to be differentiated from benign ones. However, it is well known that it is extremely difficult to differentiate benign from malignant CBD strictures on the basis of radiologic features alone, especially in patients with only a focal CBD stricture. Moreover, given the now widespread use of imaging studies in patients with jaundice, early detection of biliary stenosis is possible, and it is not rare to encounter patients with a focal CBD stricture that is difficult to characterize. To our knowledge, no reports are available concerning the usefulness of helical CT for differentiating malignant from benign CBD strictures in a large population. Thus, the purpose of our study was to evaluate retrospectively the use of multiphasic helical CT for differentiating malignant from benign CBD strictures in patients with only a focal CBD stricture and to determine predictors for this differentiation. MATERIALS AND METHODS TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferencesPatients and Reference Standard DiagnosesA computerized search by one of the authors (S.H.C.) of our hospitals radiology and pathology files for the 4-year period from 1999 to 2002 revealed that 180 patients were suspected of having a CBD stricture. One hundred thirty patients were excluded from the analysis by that author for one of the following reasons: (a) a CT study had not been performed; (b) obstructive jaundice was not confirmed by means of laboratory testing and direct cholangiography; (c) the patient had undergone biliary interventional procedures, such as biliary stent insertion, endoscopic biliary drainage, or percutaneous transhepatic biliary drainage, before the CT study; (d) the review of CT images or medical records revealed nodal or extrabiliary metastasis, acute or chronic pancreatitis, or biliary stones; or (e) pathology reports revealed exophytic or polypoid cholangiocarcinoma of the CBD. The remaining 50 patients with a malignant (n = 32) or a benign (n = 18) CBD stricture were included in our retrospective study. Their mean age was 61.6 years (age range, 3587 years). The patients included 35 men (age range, 3587 years; mean age, 61.8 years) and 15 women (age range, 3874 years; mean age, 61.1 years). According to our hospitals institutional review board, we were not required to have formal approval or informed patient consent for the limited and anonymous review of patient data required for this study. All malignant CBD strictures were infiltrative cholangiocarcinomas. The diagnosis of all malignant CBD strictures and five of the 18 benign strictures was confirmed by reviewing patients surgical and pathology records. All patients with a malignant stricture underwent a Whipple operation (n = 18) or pylorus-preserving pancreaticoduodenectomy (n = 14), one patient with a benign stricture underwent a Whipple operation, and four patients with benign strictures underwent CBD resection only. In 13 of 18 patients with benign CBD strictures, the diagnosis was confirmed by means of clinical features. The diagnostic criterion for a benign stricture was a stable condition during a follow-up period of more than 1 year after the endoscopic CBD biopsy (7,8); the median length of follow-up was 13.7 months (range, 1220 months). CT Scan AcquisitionIn the 50 patients included in our study, CT examinations had been performed at our institution with a Somatom Plus-4 scanner (Siemens, Erlangen, Germany) (n = 24), a HiSpeed Advantage scanner (General Electric Medical Systems, Milwaukee, Wis) (n = 11), or an MX8000 fourdetector row scanner (Philips Medical Systems, Cleveland, Ohio) (n = 15). Each patient received 120 mL of a nonionic contrast material (iopromide, Ultravist 370; Schering Korea, Seoul, Korea) through an 18-gauge angiographic catheter inserted into a forearm vein. CT scans were routinely obtained with the patient in a supine position during full inspiration. The contrast material was injected at a rate of 3 mL/sec with an automatic power injector. Helical CT was performed with a singledetector row scanner at the following parameters: detector collimation, 5 mm; table pitch, 1:1; and reconstruction intervals, 5 mm. When the MX8000 scanner was used, the parameters were as follows: detector collimation, 2.5 mm; table speed, 20 mm/sec; section thickness, 3.2 mm; and reconstruction interval, 1.6 mm. Multiphasic helical CT scans were obtained 30 seconds (hepatic arterial phase) and 70 seconds (portal venous phase) after initiation of the contrast material injection. CT Scan AnalysisThe CT scans acquired in the 50 patients were reviewed in consensus by two abdominal radiologists (J.M.L., 14 years of experience; S.H.K., 7 years of experience) who had no knowledge of the final radiologic or pathologic findings. All CT scans were reviewed on a picture archiving and communication system workstation (Marotech, Seoul, Korea). During analysis of the CT features, cases of malignant and benign strictures were randomly intermixed. The CBD was considered to be involved when the two radiologists encountered a narrow duct followed by a dilated proximal duct. CT findings were interpreted with regard to wall thickness, the location and length of the involved CBD, and the enhancement pattern of the involved CBD wall during hepatic arterial and portal venous phases. These were compared with the attenuation of the normal CBD wall, the maximum CBD diameter proximal to the site of the involved CBD, and the presence or absence of pancreatic duct dilatation, air in the biliary tree, and cholecystectomy. The maximum transverse thickness of the involved CBD wall was recorded, and a thick CBD wall was defined as wall thickening of more than 1.5 mm (9). The location of the involved duct was described as intrapancreatic, suprapancreatic, or supra- and intrapancreatic. The involved CBD length was calculated by observing a number of CT scans. CT numbers were obtained by means of region-of-interest cursors placed on all wall portions of the CBD lesion and on the dilated upstream CBD wall (J.M.L. or S.H.K.). These cursors were carefully placed to encompass as much of the CBD wall as possible and to avoid adjacent structures. Region-of-interest measurements for lesions were obtained for both hepatic arterial phase and portal venous phase CT scans by either author. In all cases, scans were later reviewed in consensus to ensure that region-of-interest measurements were obtained for each CBD lesion and for the upstream CBD. The mean CT number was recorded for each measurement obtained. In all cases, the individual measurements were very close to the mean. A difference of more than 10 HU in mean CT numbers between the involved CBD wall and the upstream CBD wall was considered meaningful. Finally, the CT number of the involved CBD wall was compared with that of the normal upstream CBD wall, and the enhancement pattern of the involved CBD wall was classified as hyperenhanced, isoenhanced, or hypoenhanced in comparison with that of the normal CBD wall. Histopathologic FindingsThe histopathologic findings in patients with malignant CBD strictures were collected by one radiologist (S.H.C.) who reviewed the pathology reports. The length of the involved duct was recorded, and its location was also classified as intrapancreatic, suprapancreatic, or supra- and intrapancreatic on the basis of the histopathologic findings. Statistical AnalysisThe Kolmogorov-Smirnov test was used to determine whether values were normally distributed. Statistical differences in the CT features of malignant and benign strictures were analyzed with the 2 test, the Fisher exact test, or the Mann-Whitney U test. For malignant strictures, the paired Student t test and the McNemar test were used to assess differences between CT and histopathologic findings, and agreements between CT and histopathologic findings were assessed by calculating the statistic. A multivariable stepwise logistic regression model was used to determine the best predictors of a differential diagnosis between benign and malignant strictures (10). Thereafter, the probability of benign or malignant obstruction was calculated for each patient. With these data, we determined the diagnostic performance of CT for differentiating benign from malignant obstruction. Significant differences were defined as those with P values less than .05. All statistical analyses were performed with the SPSS software package (version 10.0; SPSS, Chicago, Ill). RESULTS TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferencesUnivariate AnalysisTable 1 summarizes the different CT features observed in patients with a malignant or a benign CBD stricture. The involved segments of malignant CBD strictures were significantly longer (17.9 mm 6.6 mean standard deviation) than those of benign strictures (8.9 mm 6.8) (P 1.5 mm) in 26 of 32 patients with a malignant stricture and in three of 18 patients with a benign stricture, and this difference was also significant (P .0001) (Figs 1, 2). View this table:in this windowin a new window TABLE 1. CT Findings in Malignant and Benign CBD Strictures View larger version (146K):in this windowin a new windowDownload PPT slide Figure 1a. Contrast materialenhanced transverse CT scans show infiltrative cholangiocarcinoma involving the supra- and intrapancreatic portion of the CBD in a 54-year-old man. (a) Hepatic arterial phase scan shows a dilated CBD with wall thickening (large arrow). The thickened wall is hyperattenuating relative to the wall of the normal CBD. Dilated intrahepatic ducts (small arrows) are also seen. (b) Portal venous phase scan shows progressive enhancement of the thickened CBD wall (large arrow) compared with the hepatic arterial phase scan in a. Dilated intrahepatic ducts (small arrows) are seen. (c) Hepatic arterial phase scan obtained at the level of the pancreas shows thickening of the entire ductal wall (arrows) in the intrapancreatic portion of the CBD. The thickened wall shows hyperenhancement. (d) Portal venous phase scan shows thickening of the entire ductal wall (arrows) in the intrapancreatic portion of the CBD. The thickened wall shows more enhancement here than during the hepatic arterial phase. View larger version (151K):in this windowin a new windowDownload PPT slide Figure 1b. Contrast materialenhanced transverse CT scans show infiltrative cholangiocarcinoma involving the supra- and intrapancreatic portion of the CBD in a 54-year-old man. (a) Hepatic arterial phase scan shows a dilated CBD with wall thickening (large arrow). The thickened wall is hyperattenuating relative to the wall of the normal CBD. Dilated intrahepatic ducts (small arrows) are also seen. (b) Portal venous phase scan shows progressive enhancement of the thickened CBD wall (large arrow) compared with the hepatic arterial phase scan in a. Dilated intrahepatic ducts (small arrows) are seen. (c) Hepatic arterial phase scan obtained at the level of the pancreas shows thickening of the entire ductal wall (arrows) in the intrapancreatic portion of the CBD. The thickened wall shows hyperenhancement. (d) Portal venous phase scan shows thickening of the entire ductal wall (arrows) in the intrapancreatic portion of the CBD. The thickened wall shows more enhancement here than during the hepatic arterial phase. View larger version (148K):in this windowin a new windowDownload PPT slide Figure 1c. Contrast materialenhanced transverse CT scans show infiltrative cholangiocarcinoma involving the supra- and intrapancreatic portion of the CBD in a 54-year-old man. (a) Hepatic arterial phase scan shows a dilated CBD with wall thickening (large arrow). The thickened wall is hyperattenuating relative to the wall of the normal CBD. Dilated intrahepatic ducts (small arrows) are also seen. (b) Portal venous phase scan shows progressive enhancement of the thickened CBD wall (large arrow) compared with the hepatic arterial phase scan in a. Dilated intrahepatic ducts (small arrows) are seen. (c) Hepatic arterial phase scan obtained at the level of the pancreas shows thickening of the entire ductal wall (arrows) in the intrapancreatic portion of the CBD. The thickened wall shows hyperenhancement. (d) Portal venous phase scan shows thickening of the entire ductal wall (arrows) in the intrapancreatic portion of the CBD. The thickened wall shows more enhancement here than during the hepatic arterial phase. View larger version (151K):in this windowin a new windowDownload PPT slide Figure 1d. Contrast materialenhanced transverse CT scans show infiltrative cholangiocarcinoma involving the supra- and intrapancreatic portion of the CBD in a 54-year-old man. (a) Hepatic arterial phase scan shows a dilated CBD with wall thickening (large arrow). The thickened wall is hyperattenuating relative to the wall of the normal CBD. Dilated intrahepatic ducts (small arrows) are also seen. (b) Portal venous phase scan shows progressive enhancement of the thickened CBD wall (large arrow) compared with the hepatic arterial phase scan in a. Dilated intrahepatic ducts (small arrows) are seen. (c) Hepatic arterial phase scan obtained at the level of the pancreas shows thickening of the enti