Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print February 27, 2004, 10.1148/radiol.2311030017
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2311030017v1
231/1/101    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ward, J.
Right arrow Articles by Robinson, P. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ward, J.
Right arrow Articles by Robinson, P. J.
(Radiology 2004;231:101-108.)
© RSNA, 2004

Gastrointestinal Imaging

Bile Duct Strictures after Hepatobiliary Surgery: Assessment with MR Cholangiography1

Janice Ward, MSc, DCR, Maria B. Sheridan, BA, MB, MRCP, FRCR, J. Ashley Guthrie, BA, MB, MRCP, FRCR, Mervyn H. Davies, MD, FRCP, Charles E. Millson, MD, MRCP, J. Peter A. Lodge, MD, FRCS, Stephen G. Pollard, MA, MS, FRCS, Kondragunta R. Prasad, MB, MS, FRCS, Giles J. Toogood, MA, DM, FRCS and Philip J. Robinson, MB, BS, FRCP, FRCR

1 From the MRI Department, Clinical Radiology (J.W., M.B.S., J.A.G., P.J.R.) and Hepatobiliary and Transplant Unit (M.H.D., C.E.M., J.P.A.L., S.G.P., K.R.P., G.J.T.), St James’s University Hospital, Beckett St, Leeds LS9 7TF, England. Received January 20, 2003; revision requested April 11; final revision received August 15; accepted September 29. Address correspondence to J.W. (e-mail: janice.ward@leedsth.nhs.uk).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To establish the accuracy of magnetic resonance (MR) cholangiography for diagnosis of postsurgical bile duct strictures.

MATERIALS AND METHODS: Sixty-seven patients suspected of having bile duct strictures after liver transplantation (n = 54), cholecystectomy (n = 8), hepatic resection (n = 4), or pancreaticoduodenectomy (n = 1) underwent MR cholangiography. Thick-slab single-shot fast spin-echo (repetition time msec/echo time msec, 4,500/940) imaging was performed in the coronal through sagittal planes with rotation in 10° increments, and contiguous thin-section images were obtained in the transverse and the optimal coronal oblique planes by using half-Fourier rapid acquisition with relaxation enhancement (1,900/96). Three blinded observers independently reviewed the MR images and recorded diagnostic features including presence of biliary stricture by using a five-point confidence scale. Receiver operating characteristic analysis was used to measure the accuracy of MR cholangiography. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. Final diagnosis was established at surgery (n = 29) and direct cholangiography (23 of 29) or at direct cholangiography, liver biopsy, and/or serial liver function tests (n = 38).

RESULTS: Thirty-three of 67 patients had strictures confirmed with the reference standard. MR cholangiography enabled correct diagnosis and depicted the site of strictures in all cases. Findings of stricture at MR cholangiography were false-positive in five patients with moderate duct dilatation and caliber change at the level of the anastomosis. Mean accuracy, sensitivity, specificity, PPV, and NPV were 94%, 97%, 74%, 86%, and 96%, respectively.

CONCLUSION: MR cholangiography is as sensitive as direct cholangiography for the assessment of bile duct strictures after hepatobiliary surgery but may lead to overestimation of the importance of duct dilatation and caliber change.

© RSNA, 2004

Index terms: Bile ducts, MR, 76.121419 • Bile ducts, stenosis or obstruction, 76.453, 76.458 • Endoscopic retrograde cholangiopancreatography (ERCP), 76.1222 • Magnetic resonance (MR), cholangiopancreatography, 76.121419


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
With improvements in surgical techniques in recent years, there has been a substantial reduction in the incidence of biliary complications of liver transplantation and hepatobiliary surgery. Nevertheless, bile duct injuries are a serious problem and a major cause of morbidity. Bile duct injury occurs at cholecystectomy (whether laparoscopic or open) in 0.1%–1.3% of patients (1–6). Although the incidence is low, the problem is important because it involves a large group of patients; more than 500,000 cholecystectomies are performed annually in North America, and most bile duct injuries occur at cholecystectomy. In addition, bile duct injuries account for the highest number of complications of hepatobiliary surgeries overall. Early complications may include bile duct injury caused by mistakenly placed clips, erroneous cutting of bile ducts based on misinterpretation of biliary anatomy, periductal bile leakage that causes edema, fibrosis and secondary stricturing, and ischemia due to injury to the right hepatic artery. Bile duct strictures are the most common of the late complications and can develop a few months or many years after surgery (4–6).

Between 10% and 33% of patients develop bile duct complications after liver transplantation, and approximately 1% of grafts are lost as a result of technical failure (7–11). Using standardized transplantation techniques, Greif and colleagues (11) from the University of Pittsburgh found serious bile duct complications in 11.5% of patients with liver transplants, with biliary stricture being the most common complication. Although endoscopic or percutaneous treatments are appropriate in selected patients (7), surgical revision is the definitive treatment for bile duct strictures (11). Surgery should be performed soon after the onset of symptoms, to avoid secondary cholangitis. Cholangitis may lead to further stricturing, and, in liver transplant patients, may threaten graft survival. A timely and accurate diagnosis is imperative and requires visualization of the entire biliary tree.

Endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTHC) have traditionally been used to diagnose bile duct strictures, but both methods are invasive, time consuming, and associated with low but important morbidity of 5%–7% and mortality of less than 1% (12). Moreover, depending on the experience of the endoscopist, ERCP may be unsuccessful in approximately 18% of patients (13,14). Magnetic resonance (MR) cholangiography has been described as a noninvasive alternative to direct cholangiography with ERCP or PTHC for diagnosis of bile duct disorders, and several authors have reported high accuracies for MR cholangiography in the detection of bile duct strictures (13–24). Most previous studies, however, were limited by small study populations (15,16,18–21,23,24) or incomplete verification of pathologic findings (15,18,19,21–23), and further validation of these initial results has been recommended (19). Thus, the purpose of our study was to establish the accuracy of MR cholangiography for the assessment of postsurgical bile duct strictures.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Patients
Between February 2000 and December 2001, 90 consecutive patients suspected of having biliary strictures after hepatobiliary surgery were referred for MR cholangiography. Of the initial 90, one claustrophobic patient, one patient with breath-holding difficulties, and 21 patients without direct cholangiographic or surgical correlation were excluded from the study. The final study group comprised 67 patients with an age range of 29–72 years (mean age, 53 years); 35 of the 67 were men (age range, 29–66 years; mean age, 51 years), and 32 were women (age range, 39–72 years; mean age, 55 years). All patients had abnormal liver function test results, and 12 had a clinical history of cholangitis. The surgical procedures performed prior to MR cholangiography were liver transplantation in 54 patients, cholecystectomy in eight, hepatic resection in four, and pancreaticoduodenectomy in one. The underlying causes of liver disease in the patients with liver transplants were viral hepatitis C (n = 15), hepatitis B (n = 3), primary biliary cirrhosis (n = 14), alcoholic liver disease (n = 11), autoimmune hepatitis (n = 3), cryptogenic cirrhosis (n = 3), primary sclerosing cholangitis (n = 2), Caroli disease (n = 1), paracetamol overdose (n = 1), and metastatic liver disease (n = 1). Local ethics committee approval was granted, and written informed consent was obtained from each patient prior to entry into the study.

MR Imaging
MR imaging was performed with a 1.5-T clinical imager (Symphony; Siemens, Erlangen, Germany) by using a 30-mT/m gradient and a phased-array body coil. After localization of the biliary tree by using a true fast imaging with steady-state precession sequence (repetition time msec/echo time msec, 5.57/2.79; flip angle, 80°), thick-slab single-shot fast spin-echo (SE) (4,500/940; echo train length, 240; matrix, 360 x 512; collimation, 50 mm; acquisition time, 5 seconds) MR cholangiographic images were acquired in the coronal through sagittal planes with rotation in 10° increments around a point immediately anterior to the splenic-portal venous confluence. Contiguous thin-section MR cholangiographic images also were obtained by using half-Fourier rapid acquisition with relaxation enhancement (1,900/96–110; echo train length, 128; matrix, 218 x 512; section thickness, 4 mm; 11 sections acquired; acquisition time, 20 seconds) applied in the optimal coronal oblique plane, which was selected on the basis of thick-slab images, and in the transverse plane at the level of the anastomosis to ensure that any abnormality shown on the initial thick-slab images was included in the field of view. The field of view was 28–35 cm, depending on patient size. Parallel saturation bands were placed immediately adjacent to the region of interest, and fat suppression was applied for all sequences. All images were obtained during arrested expiration. All patients had fasted for 4 hours prior to the procedure, and we did not administer an oral contrast agent prior to imaging. In all cases, the images were of diagnostic quality.

Image Analysis
Images from MR cholangiography were viewed independently by three observers (J.W., J.A.G., P.J.R.), each with at least 4 years of experience, who were informed only of the type of biliary anastomosis in each patient and otherwise were blinded to the patient’s clinical history, as well as to the results of other imaging studies and the interpretations by other observers. Visualization of the anastomosis was recorded, and depiction of the native or donor extrahepatic bile duct and the first-, second-, and third-order intrahepatic bile ducts was characterized as excellent (complete delineation), good (>=90% delineation), fair (<90% delineation), or absent (nondepiction). The presence of biliary stricture was assessed on the basis of a five-point confidence scale according to which a score of 1 equaled definite absence, 2 equaled probable absence, 3 equaled possible presence, 4 equaled probable presence, and 5 equaled definite presence of stricture. Stricture was defined as narrowing of the bile duct lumen that was judged by the observer to be sufficient to impair bile flow and to explain abnormal biochemistry (ie, an elevation in the measured blood serum levels of liver enzymes and bilirubin). The following diagnostic features associated with each biliary stricture were recorded: intra- and extrahepatic bile duct dilatation, defined as a diameter exceeding 2 mm and 7 mm, respectively; presence and level of ductal caliber change; nonvisualization of part of the duct, or anastomosis with clear visualization of the duct on either side; and the presence of stones or sludge in the ducts. The observers also were asked to record whether strictures were single or multiple. Multiple strictures in the same patient were not scored separately.

Direct cholangiography and analysis.—Sixty-one patients underwent both MR cholangiography and direct cholangiography. The static ERCP and PTHC images were evaluated separately for the presence of bile duct stricture in consensus by three different experienced observers (M.B.S., M.H.D., C.E.M.) with a minimum of 7 years of experience each. Six patients with surgically confirmed strictures did not undergo direct cholangiography. In all six, liver function test results were grossly abnormal and all of the diagnostic features of stricture were present on MR cholangiographic images. Since surgery was considered the treatment of choice in these patients, we considered it inappropriate to perform further invasive imaging prior to surgery. Direct cholangiography was performed after puncture of an intrahepatic duct by using a 22-gauge needle with either ultrasonographic or fluoroscopic guidance (in PTHC) or after endoscopic cannulation of the recipient common bile duct via the ampulla of Vater (in ERCP).

Reference standard.—The presence or absence of stricture was established on the basis of findings at surgery in 29 patients (23 of whom also underwent direct cholangiography) and at direct cholangiography, liver biopsy, and/or serial biochemistry analysis (ie, outcome) in 38 patients. Serial biochemistry analysis (ie, liver function tests), which comprised measurement of serum levels of alkaline phosphatase, alanine transaminase, and bilirubin, was used to measure outcome on the basis of blood serum levels that were elevated at presentation and either remained elevated or returned to normal levels over time. MR cholangiography and direct cholangiography were performed within 24 hours of one another in 30 patients, within 2–7 days in 19 patients, within 8–14 days in five patients, and within 15–28 days in three patients. In four patients, there was a delay of more than 30 days between the two examinations, but, because the results were concordant, the patients were included in our final analysis. MR cholangiography and surgery were performed within 24 hours of one another in two patients, within 2–7 days in four patients, within 8–14 days in five patients, within 15–28 days in three patients, within 29–90 days in 12 patients, and within 91 days or more in three patients.

Statistical analysis.—The accuracy of MR cholangiography was measured with receiver operating characteristic (ROC) analysis by using the five-point rating scale. The area under the ROC curve was calculated, and 95% CIs were assigned by using the method of McNeil and Hanley (25) for each observer. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) also were calculated. For each observer, 95% CIs also were calculated for each of these statistics (26). A score of 1 or 2 was considered negative for stricture, whereas a score of 3 or higher was considered positive. Interobserver variability was assessed with {kappa} scores ({kappa} <= 0.40 indicated poor correlation, 0.41–0.75 indicated good correlation, and >= 0.76 indicated excellent correlation).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Forty-seven of 48 duct-to-duct anastomoses and 11 of 11 biliary-enteric anastomoses were visualized with MR cholangiography. An anastomosis between the gallbladder and the common bile duct in one patient was also visible at MR cholangiography. In the one patient in whom the anastomosis was not detected, there was good depiction of the native common bile duct, but the donor common bile duct was collapsed and surrounded by periportal edema. Among the seven patients who underwent cholecystectomy (n = 5) and hepatic resection (n = 2) without surgery of the common bile duct, the entire length of the duct was well depicted in six. In one patient with an extensive bile duct injury confirmed at surgery, there was nondepiction of the common hepatic duct just below the liver hilum. Grades for depiction of the intra- and extrahepatic ducts are shown in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Grades for Depiction of Intra- and Extrahepatic Bile Ducts at MR Cholangiography

 
Strictures
Thirty-three patients had strictures confirmed at surgery and direct cholangiography (n = 20), at surgery alone (n = 5), or at direct cholangiography with dynamic imaging and at serial biochemistry analysis (n = 8). These strictures occurred after liver transplantation in 24 patients, cholecystectomy in five, hepatic resection in three, and pancreaticoduodenectomy in one. The strictures involved only the anastomosis in 24 patients (18 duct-to-duct and six biliary-enteric anastomoses). In four patients with duct-to-duct anastomosis, strictures were solitary and distal to the anastomosis in one patient, solitary and proximal to the anastomosis in one, and multiple (ie, involving more than one duct) and proximal to the anastomosis at the liver hilum in two. All three observers correctly documented multiple strictures in these two patients. Two patients had strictures caused by extrinsic compression due to lymphoma after liver transplantation in one and to recurrent adenocarcinoma after hepatic resection in the other. In both patients, MR cholangiography depicted not only the site of obstruction but also the cause. Of the five patients with intact (nonanastomotic) bile ducts and strictures, four had strictures at the liver hilum (after injury at cholecystectomy in two and after injury at hepatic resection in two), and one had an intrahepatic stricture at the level of the origin of the right intrahepatic duct after injury at cholecystectomy (Fig 1). MR cholangiography correctly depicted the presence and site of the stricture in all cases.



View larger version (163K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Intrahepatic bile duct injury after laparoscopic cholecystectomy in a 73-year-old woman. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image shows dilatation of the right-sided ducts and nonvisualization of the ductal anatomy at the point where the right-sided ducts converge (arrow). The high-signal-intensity foci (arrowheads) are small cholangitic abscesses. (b) Corresponding ERCP image shows normal left-sided ducts; right-sided ducts are not depicted because of a tight stricture. Note the surgical clip at the point of stricture (arrow).

 


View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Intrahepatic bile duct injury after laparoscopic cholecystectomy in a 73-year-old woman. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image shows dilatation of the right-sided ducts and nonvisualization of the ductal anatomy at the point where the right-sided ducts converge (arrow). The high-signal-intensity foci (arrowheads) are small cholangitic abscesses. (b) Corresponding ERCP image shows normal left-sided ducts; right-sided ducts are not depicted because of a tight stricture. Note the surgical clip at the point of stricture (arrow).

 
The areas under the ROC curve and the 95% CIs for MR cholangiography with each observer are shown in Table 2. Sensitivity, specificity, and 95% CIs for MR cholangiography with each observer are shown in Table 3. PPV, NPV, and 95% CIs for MR cholangiography with each observer are shown in Table 4. Twenty-six of 33 strictures were associated with duct dilatation and caliber change (Fig 2); in 20 of these, nondepiction of part of the duct was recorded (Fig 3). Duct dilatation was present in all six patients who had strictures of a biliary-enteric anastomosis (Fig 4). In one patient with a duct-to-duct anastomosis and a stricture confirmed at subsequent surgery, the stricture was correctly diagnosed at MR cholangiography on the basis of a large stone located immediately proximal to the anastomosis (Fig 5). Bile duct stones proximal to strictures were identified at MR cholangiography in 15 patients and confirmed at direct cholangiography in six and at surgery in nine. In two patients, stones identified at MR cholangiography by all three observers were not confirmed at surgery (n = 1) or at PTHC (n = 1).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Area under the ROC Curve with 95% CIs for Three Observers

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Sensitivity and Specificity with 95% CIs for Three Observers

 

View this table:
[in this window]
[in a new window]

  		TABLE 4. PPV and NPV with 95% CIs for Three Observers

 


View larger version (165K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Surgically confirmed anastomotic stricture after orthotopic liver transplantation in a 54-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image shows a caliber change at the level of the anastomosis (arrow) and dilated intrahepatic ducts consistent with anastomotic stricture. (b) Corresponding ERCP image depicts no flow of contrast material across the anastomosis (arrow). A tight anastomotic stricture was confirmed at subsequent surgery.

 


View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Surgically confirmed anastomotic stricture after orthotopic liver transplantation in a 54-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image shows a caliber change at the level of the anastomosis (arrow) and dilated intrahepatic ducts consistent with anastomotic stricture. (b) Corresponding ERCP image depicts no flow of contrast material across the anastomosis (arrow). A tight anastomotic stricture was confirmed at subsequent surgery.

 


View larger version (178K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Surgically confirmed common bile duct stricture after laparoscopic cholecystectomy in a 43-year-old man. Coronal oblique single-shot fast SE (4,500/940) MR image shows intrahepatic duct dilatation and caliber change but does not depict part of the common bile duct (arrow).

 


View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Surgically confirmed biliary-enteric anastomotic stricture after orthotopic liver transplantation in a 56-year-old woman. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image clearly depicts the Roux-en-Y loop but not the point where it joins the ducts (arrow). Note the dilatation of the intrahepatic ducts. (b) Corresponding PTHC image shows that only a small amount of contrast material has passed beyond the stricture and entered the Roux-en-Y loop (arrow).

 


View larger version (186K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Surgically confirmed biliary-enteric anastomotic stricture after orthotopic liver transplantation in a 56-year-old woman. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image clearly depicts the Roux-en-Y loop but not the point where it joins the ducts (arrow). Note the dilatation of the intrahepatic ducts. (b) Corresponding PTHC image shows that only a small amount of contrast material has passed beyond the stricture and entered the Roux-en-Y loop (arrow).

 


View larger version (178K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a. Surgically confirmed anastomotic stricture after orthotopic liver transplantation in a 54-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image depicts minimal intrahepatic duct dilatation and a stone (long arrow) immediately proximal to the anastomosis (short arrow). (b) Corresponding ERCP image also shows the stone (arrow). (c) Transverse and (d) coronal oblique views from thin-section half-Fourier rapid acquisition with relaxation enhancement (1,900/100) MR cholangiography show the stone (arrow) more clearly than does a.

 


View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b. Surgically confirmed anastomotic stricture after orthotopic liver transplantation in a 54-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image depicts minimal intrahepatic duct dilatation and a stone (long arrow) immediately proximal to the anastomosis (short arrow). (b) Corresponding ERCP image also shows the stone (arrow). (c) Transverse and (d) coronal oblique views from thin-section half-Fourier rapid acquisition with relaxation enhancement (1,900/100) MR cholangiography show the stone (arrow) more clearly than does a.

 


View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5c. Surgically confirmed anastomotic stricture after orthotopic liver transplantation in a 54-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image depicts minimal intrahepatic duct dilatation and a stone (long arrow) immediately proximal to the anastomosis (short arrow). (b) Corresponding ERCP image also shows the stone (arrow). (c) Transverse and (d) coronal oblique views from thin-section half-Fourier rapid acquisition with relaxation enhancement (1,900/100) MR cholangiography show the stone (arrow) more clearly than does a.

 


View larger version (200K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5d. Surgically confirmed anastomotic stricture after orthotopic liver transplantation in a 54-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) MR cholangiographic image depicts minimal intrahepatic duct dilatation and a stone (long arrow) immediately proximal to the anastomosis (short arrow). (b) Corresponding ERCP image also shows the stone (arrow). (c) Transverse and (d) coronal oblique views from thin-section half-Fourier rapid acquisition with relaxation enhancement (1,900/100) MR cholangiography show the stone (arrow) more clearly than does a.

 
False-Positive Findings at MR Cholangiography
At diagnostic confidence scores of 3–5, strictures were diagnosed at MR cholangiography by at least two of three observers in six patients (after transplantation in five patients and cholecystectomy in one) with moderate duct dilatation and caliber change who were shown not to have strictures according to the standard of reference (surgery alone [n = 1], surgery and direct cholangiography [n = 2], direct cholangiography and outcome [n = 3]) (Fig 6). In one of these six patients, stones in a gallbladder conduit that were diagnosed at MR and confirmed at subsequent surgery were causing obstruction at the level of the anastomosis, but there was no associated stricture. In one patient, ERCP showed sludge in the bile duct, which was successfully treated with sphincterotomy; and in one patient with a bile leak shown at ERCP and confirmed at surgery, smooth tapering of the donor common bile duct immediately proximal to the anastomosis depicted at MR cholangiography was probably the result of extrinsic compression by a local fluid collection. In one patient, no major stricture was seen at ERCP, but a stent was inserted into the common bile duct because of minor caliber change. Although a slight decrease in liver enzyme level occurred after stent insertion, liver function test results continued to improve after stent removal. Liver function test results also returned to normal after steroid treatment in one patient in whom a stent was not inserted at ERCP. Outcome indicated that no substantial stricture was present in these two patients. In one further patient who had undergone open cholecystectomy, stones found in the common bile duct and the right and left intrahepatic ducts at surgery were treated with choledochoduodenostomy, after which liver function test results returned to normal.



View larger version (183K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a. Anastomotic stricture overestimated at MR cholangiography after orthotopic liver transplantation in a 47-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) image clearly shows minimal duct dilatation and caliber change at the level of the anastomosis (arrow). (b) Corresponding ERCP image, in which the discrepancy in the size of the native and donor ducts is less pronounced because the biliary system is distended with contrast material, which is seen to flow freely across the anastomosis (arrow). A stent was inserted because of minor caliber change, but the patient continued to improve after the stent was removed.

 


View larger version (142K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b. Anastomotic stricture overestimated at MR cholangiography after orthotopic liver transplantation in a 47-year-old man. (a) Coronal oblique single-shot fast SE (4,500/940) image clearly shows minimal duct dilatation and caliber change at the level of the anastomosis (arrow). (b) Corresponding ERCP image, in which the discrepancy in the size of the native and donor ducts is less pronounced because the biliary system is distended with contrast material, which is seen to flow freely across the anastomosis (arrow). A stent was inserted because of minor caliber change, but the patient continued to improve after the stent was removed.

 
False-Negative Findings at MR Cholangiography
At diagnostic confidence scores of 3–5, although two of the three observers correctly diagnosed all 33 strictures, one observer underestimated three strictures in three patients at MR cholangiography. In one of these patients, a stone immediately proximal to the anastomosis, which was the only diagnostic feature of stricture, was not identified. In one patient with relatively mild intrahepatic duct dilatation and a stricture at the liver hilum, nondepiction of the bifurcation was attributed to an image artifact caused by a prominent portal vein. In another patient, stones at the distal end of the common bile duct and just above the anastomosis were considered to be more in keeping with choledocholithiasis than with biliary stricture.

Findings at ERCP and PTHC
The static images obtained with direct cholangiography discorded with those obtained with the reference standard in only one of 63 patients. On static ERCP images, a stricture was underestimated in this patient, whose liver enzyme levels returned to normal after stent placement for a possible stricture demonstrated at dynamic ERCP. The stricture was correctly diagnosed at MR cholangiography on the basis of duct dilatation and caliber change, but static and dynamic ERCP images showed only minor narrowing, within normal limits, at the anastomosis.

Interobserver Agreement
{kappa} values for interobserver agreement for the diagnosis of a stricture were good to excellent (Table 5). Although all three observers achieved accuracy of more than 90%, interobserver agreement was not as good because of the different false-positive and false-negative findings by each observer in the different cases.


View this table:
[in this window]
[in a new window]

  		TABLE 5. Interobserver Agreement for Presence of Bile Duct Strictures at MR Cholangiography

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
REFERENCES
 
In previous studies, MR cholangiography has been proved extremely accurate for depiction of postsurgical bile duct strictures, with reported accuracies of 100% (15,16,18,19,21). These studies, however, were disadvantaged by small patient populations (15,16,18,19,21) or by the use of direct cholangiography as the standard of reference (15,18,19,21). Direct cholangiography is currently considered the reference standard for diagnosis of biliary strictures, but it has several limitations. In liver transplant patients, hepatic dysfunction may have a variety of causes, so there may be uncertainty regarding the importance of minimal duct dilatation and slight narrowing at the anastomosis. Moreover, the injection of contrast material may lead to low-grade strictures being overestimated because of overdistention of ducts; and because contrast material cannot pass beyond a high-grade stricture, additional strictures beyond this point are not detected. We attempted to address these issues in the design of our study by recruiting a larger population of patients who had undergone all types of hepatobiliary surgery and by using the more rigorous reference standard of findings at surgery or at direct cholangiography, serial biochemistry, and/or liver biopsy.

As in previous studies (16–19,21,22,27), in our study all ductal anastomoses except one were visualized with MR cholangiography, and there was complete delineation of the first-order intrahepatic ducts in the vast majority of our patients. We chose to inform the observers of the type of ductal anastomosis in each patient because this was analogous to clinical practice. In patients with a biliary-enteric anastomosis, this information also eliminated the possibility of the observer’s mistaking the redundant native bile duct for the recipient duct of a duct-to-duct anastomosis.

Almost half of our patients (33 of 67) had strictures, all of which were correctly diagnosed at MR cholangiography. Most strictures occurred in our transplant population (24 of 33 patients), and most strictures were solitary and were located at the anastomosis. Anastomotic strictures occur more frequently than nonanastomotic strictures and are usually caused by iatrogenic trauma and scar formation (28–30). Nonanastomotic strictures may be multiple and are most commonly caused by ischemic injury from prolonged cold ischemia time (31) or are secondary to hepatic artery thrombosis or stenosis (30,32). Of the four transplant patients with nonanastomotic strictures in our study, two had multiple strictures and angiographically confirmed hepatic artery occlusion. The effectiveness of direct cholangiography may be particularly limited in patients with multiple strictures because contrast material cannot pass beyond a high-grade obstruction, and ducts and strictures beyond such an obstruction cannot be visualized; although it may be possible to fill the ducts by applying greater pressure, this increases the risk of cholangitis. Prior to surgery, the biliary tree must be completely visualized. If any hepatic segment is excluded from the biliary-enteric anastomosis, inadequate drainage may lead to cholangitis and threaten graft survival. ERCP combined with PTHC, possibly with multiple punctures (16), may be necessary for complete delineation of the biliary tree, but it also exposes the patient to a high risk of cholangitis. At MR cholangiography, the ductal system proximal to the obstruction is well depicted, and the extent of biliary involvement is clearly delineated. Strictures are also a well-recognized late complication of biliary-enteric bypass surgery (8,11,15). MR cholangiography is particularly valuable in these patients because endoscopic access is rarely possible and PTHC has a high rate of complications.

Nine of our nontransplant patients had strictures. In four of the nine, stricture occurred during or after laparoscopic cholecystectomy: In one of these four patients, a mistakenly placed surgical clip resulted in segmental duct ligation. One patient presented 5 months after surgery complicated by bile leakage, one presented 8 months after surgery complicated by bile leakage and prolonged sepsis, and one presented 12 months after surgery. The other five patients developed strictures 1 month after open cholecystectomy complicated by bile leakage (n = 1), several months after extensive hepatic resection (n = 3), or several months after pancreaticoduodenectomy (n = 1). Surgery is the treatment of choice for biliary strictures after bile duct injury, and stent placement is typically successful only in the short term. MR cholangiography readily depicts these abnormalities, allows accurate planning of the definitive surgical procedure, and obviates diagnostic ERCP or PTHC.

It is not always possible to distinguish between strictures and other causes of obstruction at MR cholangiography, and this occurred in four of our patients. In all four patients, MR cholangiography clearly showed duct dilatation and caliber change, but these abnormalities were caused by bile duct stones in two patients, sludge in one patient, and extrinsic compression by a fluid collection in one patient. All of these features were correctly identified at MR cholangiography; but because bile duct stones and sludge formation may be associated with strictures, the diagnosis was particularly difficult. In practice, however, this difficulty rarely alters disease management.

Since MR cholangiography does not provide functional information and is unable to demonstrate resistance to flow, a small number of false-positive results is not surprising. The tendency to overestimate strictures in our study resulted in a mean PPV of 86%. However, with regard to the use of MR cholangiography as a first-line diagnostic test, overestimation of strictures is probably less critical than underestimation. Whereas overestimation may lead to an unnecessary referral for direct cholangiography, a false-negative diagnosis may delay appropriate therapy, which increases the risk of cholangitis and, in liver transplant patients, of graft rejection.

Our results and those of others suggest that it is unlikely that a stricture will be missed at MR cholangiography (15,16,18,19,21). We achieved a high mean NPV of 96% because at least two observers never failed to detect a stricture that was present. Our interobserver agreement was good to excellent, and each of the three observers achieved accuracy of more than 90%. Despite similar accuracies, individual sensitivities and specificities were variable because the observers were operating at different points on the ROC scale. Interobserver variation is a feature of all studies in which multiple observers are used; and when several different observations are involved in decision making, as was the case in our study, the range of interobserver variation increases (33). In other radiologic contexts, observer access to clinical data and previous images reduces such variation (34,35), so there should be less interobserver disagreement in clinical practice.

A criticism of MR cholangiography is that it tends to result in overestimation of the length of strictures because the duct immediately distal to the stricture may be collapsed (15,17,22). Careful analysis of the source images, however, should reduce such overestimation. Furthermore, while it is important to recognize this potential limitation of MR cholangiography because surgical reconstruction is the treatment of choice for bile duct stricture, only the site of the most proximal element of the stricture is relevant, and it is typically the distal extent that is overestimated at MR cholangiography. Other critical information for surgical planning is the number of strictures (because cholangitis and further strictures may occur if ducts remain isolated from the biliary-enteric anastomosis) and their location (because Roux-en-Y choledochojejunostomy is indicated in patients with low common bile duct strictures, whereas intrahepatic and hilar strictures are treated with hepaticojejunostomy). In our study, the presence and level of all strictures in all patients were correctly demonstrated with MR cholangiography.

The formation of bile duct stones is a recognized complication of liver transplantation and is often associated with stricture (36). Stones were demonstrated in 17 of our transplant patients with bile duct strictures, and in one patient with a surgically confirmed stricture, the only diagnostic feature at MR cholangiography was a stone immediately proximal to the anastomosis. Smaller calculi may be obscured on thick-slab single-shot fast SE images because of limited spatial resolution, but they were readily identified on our thin-section images and easily distinguished from pneumobilia, which is always located in the nondependent portion of the duct on transverse images.

Low-grade strictures are reported to be best depicted with direct cholangiography when the ducts are distended with contrast material (37). We also expected that the dynamic aspect of ERCP would improve the diagnosis of strictures. However, not only were all strictures correctly diagnosed at MR cholangiography in our study, but the dynamic aspect of ERCP actually led to a false-negative diagnosis in one patient. The discrepancy in size of the native and donor ducts was less pronounced at ERCP because the biliary system was distended with contrast material. Because of the MR cholangiographic findings in this patient, as well as clinical findings of jaundice and ascending cholangitis, a stent was inserted, after which the patient’s condition improved.

In an attempt to establish the added value provided by functional information obtained with direct cholangiography, we performed a separate analysis of static ERCP or PTHC images alone. Dynamic and static images were entirely concordant. Compared with our reference standard, the static images were discordant in only one patient, and it was the same patient in whom dynamic imaging also resulted in a false-negative diagnosis.

A limitation of our study is the relatively small number of patients with bile duct strictures after hepatobiliary surgery other than liver transplantation. However, our sample size compares favorably with those in earlier studies (24) and provides important validation of previous findings.

In conclusion, MR cholangiography is as sensitive as direct cholangiography for the assessment of biliary strictures. Although there is a tendency to overestimate the importance of moderate duct dilatation and minimal caliber change, our results suggest that it is highly unlikely that a biliary stricture will be missed at MR cholangiography. Our results have confirmed MR cholangiography to be an accurate and noninvasive alternative to direct cholangiography, and we now regard it as the primary imaging modality in patients suspected of having bile duct strictures.


   ACKNOWLEDGMENTS
 
The authors thank Sheila Boyes for assistance with the preparation of the manuscript and statistical results, and Dan Wilson, MSc, for statistical advice.


   FOOTNOTES
 
Abbreviations: ERCP = endoscopic retrograde cholangiopancreatography, NPV = negative predictive value, PPV = positive predictive value, PTHC = percutaneous transhepatic cholangiography, ROC = receiver operating characteristic, SE = spin echo

Author contributions: Guarantor of integrity of entire study, J.W.; study concepts and design, J.W., M.B.S., P.J.R., J.A.G.; literature research, J.W.; clinical studies, all authors; data acquisition, J.W., M.B.S., J.A.G., P.J.R., M.H.D., C.E.M.; data analysis/interpretation, J.W.; statistical analysis, J.W.; manuscript preparation, J.W.; manuscript definition of intellectual content, all authors; manuscript editing, M.B.S., P.J.R., J.A.G., M.H.D., C.E.M., J.P.A.L., S.G.P., K.R.P., G.J.T.; manuscript revision/review and final version approval, J.W.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 

  1. Roslyn JJ, Binns GS, Hughes EF, Saunders-Kirkwood K, Zinner MJ, Cates JA. Open cholecystectomy: a contemporary analysis of 42,474 patients. Ann Surg 1993; 218:129-137.[Medline]
  2. Deziel DJ, Millikan KW, Economou SG, Doolas A, Ko ST, Airan MC. Complications of laparoscopic cholecystectomy: a national survey of 4,292 hospitals and an analysis of 77,604 cases. Am J Surg 1993; 165:9-14.[CrossRef][Medline]
  3. Morgenstern L, Wong L, Berci G. Twelve hundred open cholecystectomies before the laparoscopic era: a standard for comparison. Arch Surg 1992; 127:400-403.[Abstract]
  4. Slanetz PJ, Boland GW, Mueller PR. Imaging and interventional radiology in laparoscopic injuries to the gallbladder and biliary system. Radiology 1996; 201:595-603.[Abstract/Free Full Text]
  5. Richardson MC, Bell G, Fullarton GM. Incidence and nature of bile duct injuries following laparoscopic cholecystectomy: an audit of 5913 cases. West of Scotland Laparoscopic Cholecystectomy Audit Group. Br J Surg 1996; 83:1356-1360.
  6. Andren-Sandberg A, Johansson S, Bengmark S. Accidental lesions of the common bile duct at cholecystectomy. II. Results of treatment. Ann Surg 1985; 201:452-455.
  7. Hintze RE, Alder A, Veltzke W, Abou-Redyeh H, Felix R, Neuhaus P. Endoscopic management of biliary complications after orthotopic liver transplantation. Hepatogastroenterology 1997; 44:258-262.[Medline]
  8. Colonna JO, 2nd, Shaked A, Gomes AS, et al. Biliary strictures complicating liver transplantation: incidence, pathogenesis, management and outcome. Ann Surg 1992; 216:344-352.[Medline]
  9. Theilmann L, Küppers B, Kadmon M, et al. Biliary tract strictures after orthotopic liver transplantation: diagnosis and management. Endoscopy 1994; 26:517-522.[Medline]
  10. Neuhaus P, Blumhardt G, Bechstein WO, Steffen R, Platz KP, Keck H. Technique and results of biliary reconstruction using side-to-side choledochocholedochostomy in 300 orthotopic liver transplants. Ann Surg 1994; 219:426-434.[CrossRef][Medline]
  11. Greif F, Bronsther OL, Van Thiel DH, et al. The incidence, timing and management of biliary tract complications after orthotopic liver transplantation. Ann Surg 1994; 219:40-45.[Medline]
  12. Loperfido S, Angelini G, Benedetti G, et al. Major early complications from diagnostic and therapeutic ERCP: a prospective multicenter study. Gastrointest Endosc 1998; 48:1-10.[CrossRef][Medline]
  13. Taylor AC, Little AF, Hennessy OF, Banting SW, Smith PJ, Desmond PV. Prospective assessment of magnetic resonance cholangiopancreatography for noninvasive imaging of the biliary tree. Gastrointest Endosc 2002; 55:17-22.[CrossRef][Medline]
  14. Adamek HE, Albert J, Weitz M, Breer H, Schilling D, Riemann JF. A prospective evaluation of magnetic resonance cholangiopancreatography in patients with suspected bile duct obstruction. Gut 1998; 43:680-683.[Abstract/Free Full Text]
  15. Laghi A, Pavone P, Catalano C, et al. MR cholangiography of late biliary complications after liver transplantation. AJR Am J Roentgenol 1999; 172:1541-1546.[Abstract/Free Full Text]
  16. Chaudhary A, Negi SS, Puri SK, Narang P. Comparison of magnetic resonance cholangiography and percutaneous transhepatic cholangiography in the evaluation of bile duct strictures after cholecystectomy. Br J Surg 2002; 89:433-436.[CrossRef][Medline]
  17. Boraschi P, Braccini G, Gigoni R, et al. Detection of biliary complications after orthotopic liver transplantation with MR cholangiography. Magn Reson Imaging 2001; 19:1097-1105.[CrossRef][Medline]
  18. Pavone P, Laghi A, Catalano C, et al. MR cholangiography in the examination of patients with biliary-enteric anastomoses. AJR Am J Roentgenol 1997; 169:807-811.[Abstract/Free Full Text]
  19. Fulcher AS, Turner MA. Orthotopic liver transplantation: evaluation with MR cholangiography. Radiology 1999; 211:715-722.[Abstract/Free Full Text]
  20. Khalid TR, Casillas VJ, Montalvo BM, Centeno R, Levi JU. Using MR cholangiopancreatography to evaluate iatrogenic bile duct injury. AJR Am J Roentgenol 2001; 177:1347-1352.[Abstract/Free Full Text]
  21. Meersschaut V, Mortelé KJ, Troisi R, et al. Value of MR cholangiography in the evaluation of postoperative biliary complications following orthotopic liver transplantation. Eur Radiol 2000; 10:1576-1581.[CrossRef][Medline]
  22. Tang Y, Yamashita Y, Arakawa A, et al. Pancreaticobiliary ductal system: value of half-Fourier rapid acquisition with relaxation enhancement MR cholangiopancreatography for postoperative evaluation. Radiology 2000; 215:81-88.[Abstract/Free Full Text]
  23. Hussein FM, Alsumait B, Aman S, et al. Diagnosis of choledocholithiasis and bile duct stenosis by magnetic resonance cholangiogram. Australas Radiol 2002; 46:41-46.[CrossRef][Medline]
  24. Coakley FV, Schwartz LH, Blumgart LH, Fong Y, Jarnagin WR, Panicek DM. Complex postcholecystectomy biliary disorders: preliminary experience with evaluation by means of breath-hold MR cholangiography. Radiology 1998; 209:141-146.[Abstract/Free Full Text]
  25. McNeil BJ, Hanley JA. Statistical approaches to the analysis of receiver operating characteristic (ROC) curves. Med Decis Making 1984; 4:137-150.
  26. Newcombe RG. Two sided confidence intervals for the single proportion: comparison of seven methods. Stat Med 1998; 17:857-872.[CrossRef][Medline]
  27. Fulcher AS, Turner MA, Capps GW, Zfass AM, Baker KM. Half-Fourier RARE MR cholangiopancreatography: experience in 300 subjects. Radiology 1998; 207:21-32.[Abstract/Free Full Text]
  28. Oliver JH, 3rd, Federle MP, Campbell WL, Zajko A. Imaging the hepatic transplant. Radiol Clin North Am 1991; 29:1285-1298.[Medline]
  29. Nghiem HV, Tran K, Winter TC, 3rd, et al. Imaging of complications in liver transplantation. RadioGraphics 1996; 16:825-840.[Abstract]
  30. Northover JM, Terblanche J. A new look at the arterial supply of the bile duct in man and its surgical implications. Br J Surg 1979; 66:379-384.[Medline]
  31. Sanchez-Urdazpal L, Gores GJ, Ward EM, et al. Ischemic-type biliary complications after orthotopic liver transplantation. Hepatology 1992; 16:49-53.[CrossRef][Medline]
  32. Terblanche J, Allison HF, Northover JM. An ischemic basis for biliary strictures. Surgery 1983; 94:52-57.[Medline]
  33. Robinson PJ. Radiology’s Achilles’ heel: error and variation in the interpretation of the Röntgen image. Br J Radiol 1997; 70:1085-1098.[Abstract]
  34. Aideyan UO, Berbaum K, Smith WL. Influence of prior radiologic information on the interpretation of radiographic examinations. Acad Radiol 1995; 2:205-208.[CrossRef][Medline]
  35. Doubilet P, Herman PG. Interpretation of radiographs: effect of clinical history. AJR Am J Roentgenol 1981; 137:1055-1058.[Abstract/Free Full Text]
  36. Sheng R, Ramirez CB, Zajko AB, Campbell WL. Biliary stones and sludge in liver transplant patients: a 13-year experience. Radiology 1996; 198:243-247.[Abstract/Free Full Text]
  37. Varghese JC, Farrell MA, Courtney G, et al. Role of MR cholangiopancreatography in patients with failed or inadequate ERCP. AJR Am J Roentgenol 1999; 173:1527-1533.[Abstract]



This article has been cited by other articles:


Home page
 RadioGraphicsHome page
O. A. Catalano, D. V. Sahani, S. P. Kalva, M. S. Cushing, P. F. Hahn, J. J. Brown, and R. R. Edelman
MR Imaging of the Gallbladder: A Pictorial Essay
RadioGraphics, January 1, 2008; 28(1): 135 - 155.
[Abstract] [Full Text] [PDF]

Home page
 RadioGraphicsHome page
A. H. M. Caiado, R. Blasbalg, A. S. Z. Marcelino, M. da Cunha Pinho, M. C. Chammas, C. da Costa Leite, G. G. Cerri, A. C. de Oliveira, T. Bacchella, and M. C. C. Machado
Complications of Liver Transplantation: Multimodality Imaging Approach
RadioGraphics, September 1, 2007; 27(5): 1401 - 1417.
[Abstract] [Full Text] [PDF]

Home page
 RadioGraphicsHome page
C. Hoeffel, L. Azizi, M. Lewin, V. Laurent, C. Aube, L. Arrive, and J.-M. Tubiana
Normal and Pathologic Features of the Postoperative Biliary Tract at 3D MR Cholangiopancreatography and MR Imaging
RadioGraphics, November 1, 2006; 26(6): 1603 - 1620.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2311030017v1
231/1/101    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ward, J.
Right arrow Articles by Robinson, P. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ward, J.
Right arrow Articles by Robinson, P. J.