Magnetic resonance cholangiography in patients with biliary disease: its role in primary sclerosing cholangitis

Journal of Hepatology2000; 33: 520-521 Printed in Denmark All rights reserved Munksgaard

Copyright 0 European Association fix the Study of the Liver 2000

Copenhugen

Journal of Hepatology ISSN 01&Y-8278

its Paul Angulo’,

with biliary disease: in primary sclerosing cholangitis

Dawn H. Pearce 2, C. Daniel Johnson2, Jessica J. Henry’, Bret T. Petersen’ and Keith D. Lindor’

Nicholas

E LaRusso’,

‘Division of Gastroenterology and Hepatology; und ‘Department of Diugnostic Radiology, Mayo Clinic and Foundution, Rochester, MN, USA

Background/Aim: Magnetic resonance cholangiography (MRC) is a non-invasive diagnostic procedure whose role in the management of patients with primary sclerosing cholangitis (PSC) is unclear. The aim of this study was to determine the usefulness of MRC in the evaluation of the biliary tree in patients with suspected biliary disease, and in particular, PSC. Methods: MRC and invasive cholangiography (ERCP or PTC) were both performed in 73 patients, (33 male, 40 female, mean age 56 years) with clinical and/or biochemical evidence of cholestasis. Images were interpreted by two radiologists unaware of the results of other studies. Results: Forty-two patients (58%) had benign biliary disease, including 23 patients (32%) with PSC; 9 patients (12%) had malignant biliary disease; and 22 patients (300/) had a normal biliary tree. Diagnostic quality images were obtained in 73173 (100%) of MRC, and in 70/73 (96%) of invasive cholangi-

ography (68 ERCP’s, 2 PTC’s) procedures. Using ERCP/PTC findings as the reference standard, MRC had an accuracy greater than 90% in the diagnosis of normal bile ducts, biliary dilatation, biliary obstruction, bile duct stones, and PSC. Using the final diagnosis, MRC had an overall diagnostic accuracy of 90% in the detection of biliary disease compared to 97% for invasive cholangiography. Additional diagnostic/therapeutic interventions were performed during ERCP in 73% of patients with PSC and in 43% of patients without PSC (p=O.O2). Conclusions: MRC has excellent diagnostic accuracy in the presence of biliary disease. Because of its noninvasive nature, MRC may have advantages over invasive cholangiography when diagnosis is the major goal of the procedure.

Key words: Biliary disease; Cholangiography.

See Editorial on pages 659460

E

retrograde cholangiopancreatography (ERCP) and to a lesser degree, percutaneous transhepatic cholangiography (PTC), have become important and widely available diagnostic modalities in the clinical evaluation of patients with suspected biliary disease (l-3). In particular, in patients with primary sclerosing cholangitis (PSC), invasive cholangiography (ERCP PTC) is considered the gold standard for diagnosis of this condition (4). Furthermore, PSC patients frequently undergo invasive cholangiography NDOSCOPIC

Received 14 September; revised 29 December 1999: accepted 18 February 2000

Correspondence: Keith D. Lindor, Division of Gastroenterology & Hepatology, Mayo Clinic and Foundation, Rochester MN 55905, USA. Tel: 507 284 2511. Fax: 507 284 0538. e-mail: [email protected]

520

during the course of the disease for diagnostic and/or therapeutic purposes. Although generally safe, ERCP has a complication rate varying from 3% to 8% including pancreatitis, biliary sepsis, and the risks of intubation such as bleeding, perforation and aspiration. There is limited or no opacification of ducts proximal to a complete obstruction, and routine sedation or anesthesia is required (1,2). Complications of PTC are related mainly to hepatic puncture and include pain, internal and external hemorrhage and bile leak (3). ERCP and PTC require considerable training and experience and are highly operator dependent. Magnetic resonance cholangiography (MRC) is a new application of MR imaging that provides very high contrast between the hyperintense signal of the bile and the hypointense signal of the surrounding

MR cholangiography

in PSC

Patient population

TABLE 1 Final diagnosis in the patient nonulation Final diagnosis

Number (n=73)

Benign biliary disease Primary sclerosing cholangitis Bile duct stones* Liver cysts Strictures post-OLT Small-duct PSC Duodenal diverticulum Stricture CBD post-cholecystectomy Others

42 (58%) 23 (32%) 4 (5%) 1 (1.4%) 3 (4%) 1 (1.4%) 1 (1.4%) 1 (1.4%) 8 (11%)

Malignant biliary disease Pancreas carcinoma Gastric cancer Cholangiocarcinoma** Ampulla of Vater cancer

9 (12%) 6 (8%) 1 (1.4%) 1 (1.4%) 1 (1.4%)

Normal biliary tree

22 (30%)

* Besides these four patients with bile duct stones, five patients with PSC and one patient with stricture of the CBD post-OLT also had bile duct stones (no. of patients with bile duct stones= 10). ** Besides this patient with primary cholangiocarcinoma, another patient with long-standing PSC had cholangiocarcinoma (no. of patients with cholangiocarcinoma=2).

structures. As a result, MRC can provide complete visualization of the biliary tree (5). MRC is non-invasive, has no morbidity and no mortality in properly screened patients, and does not require radiation or intravenous or oral contrast media. Previous reports about MRC have suggested a good correlation with ERCP (6-8); however, its role in the management of patients with suspected biliary disease, and in particular, in patients with PSC is still unclear. Thus, our aim was to determine the usefulness of MRC in the evaluation of the biliary tree in patients with suspected biliary disease with an emphasis on patients with PSC.

Materials and Methods This was a prospective, comparative study evaluating the accuracy of MRC compared to invasive cholangiography (ERCP or PTC) in assessing the biliary tree in patients with cholestasis.

Patients with clinical symptoms and/or results of biochemical studies consistent with biliary disease (cholestasis) and referred for ERCP were enrolled. Prior to enrollment, a complete history and physical examination and measurement of serum chemistries including serum alkaline phosphatase, aminotransferases and bilirubin as well as prothrombin time were performed. In addition, ultrasound (US) and/or CT scan of the abdomen were performed as the initial imaging study in all patients in whom biliary disease was suspected. Male or female patients, age ~18 years old were included. Usual contraindications to MR scanning such as history of cardiac pacemakers, intraocular metallic foreign bodies, cerebral aneurysm clips, caval and intravascular implants as well as claustrophobia were specifically excluded in all patients. The performance of ERCP was decided upon by the responsible physician based on diagnostic or therapeutic issues and not influenced by this protocol. Patients were identified from the GI endoscopy schedule list at the time the ERCP was scheduled. The study and MRC procedure was then explained to individual patients and their consent obtained. Serum levels of CAl9-9 were determined in all patients with PSC as part of the standard of care of these patients. Biliary brushings of dominant strictures, when present, were also done during invasive cholangiography in patients with PSC enrolled in this study. Neither serum levels of CA19-9 nor brushings were positive for bile duct cancer in any of the 22 patients with PSC without this biliary malignancy. Magnetic resonance cholangiography

All MR examinations were performed using a General Electric Signa 1.5 Tesla MR scanner. All acquisitions were performed using a phase array torso multicoil. Two acquisitions were performed in the axial and coronal planes using the single shot fast spin echo pulse sequence with a TR=infinite, a TE=90 ms, matrix size=256/256, slice thickness of 5 mm, with 0 mm skip. Each slice required 2 s. The total breath hold averaged 20 s. Multiple breath holds scans, when necessary, were utilized to cover the entire liver. Three additional acquisitions were also acquired in the coronal plane and in planes approximately 45 degrees oblique to the hepatic porta. The same single shot spin echo sequence was used as stated above, except the slice thickness was 5 cm, requiring a single 2 s breathhold. Fat saturation technique was employed. Maximum intensity projection technique was not used. MRC procedures were performed within 24 h preceding the scheduled ERCP Image interpretation

MRC and invasive cholangiography images were interpreted by two radiologists with experience in hepatic and biliary tract imaging. They were unaware of the clinical and biochemical data, and the indication and results of other imaging studies or the final diagnosis. Differences in image interpretation were resolved by consensus between the two radiologists. MRC images interpretation was done prior to invasive cholangiography interpretation. All MRC sequences were evaluated in each case for overall quality which was classified as satisfactory or unsatisfactory based on visual-

TABLE 2. Usefulness of MRC in the evaluation of the biliary tree using ERCP/PTC findings as the standard of reference (n=70)

Normal ducts Dilatation Hepatic ducts CBD Obstruction Biliary stones* PSC

Sensitivity

Specificity

Positive predictive value

Negative predictive value

Diagnostic accuracy

19/22 (86%)

46148 (96%)

19/21 (90%)

46149 (94%)

65/70 (93%)

40142 (95%) 38/41 (93%) 37/37 (100%) 5/10 (50%) 19/23 (83%)

24/28 (86%) 27129 (93%) 30/33 (91%) 59/60 (98%) 46147 (98%)

40/44 (91%) 30/40 (95%) 37/40 (92%) 5/6 (83%) 19/20 (95%)

24126 (92%) 27130 (90%) 30/30 (100%) 59164 (92%) 46150 (92%)

64/70 (91%) 65170 (93%) 67170 (96%) 64170 (91%) 65170 (93%)

* Sensitivity of MRC in the detection of bile duct stones was greater in patients without PSC (80%) compared to patients with PSC (20%).

P. Angulo et ul. ization of biliary structures and presence of artifacts. The images were evaluated for common bile duct visualization, dilatation and obstruction. Hepatic ducts were evaluated for extent of visualization and morphology including the intrahepatic and extrahepatic biliary tree. Visualization of the gallbladder and cystic duct and presence and location of biliary stones were also recorded. The level of obstruction was determined and divided into suprapancreatic, intrapancreatic, and periampullary. The same criteria were applied in the evaluation of invasive cholangiography (ERCP/PTC) using the known diameter of the endoscope as a correction reference for magnification. The visual characteristics of ampulla and additional diagnostic and/or therapeutic interventions provided concurrently during invasive cholangiography were re-

Fig. 2. ERCP (a) and MRC (b) show a focally nurroweti lower CBD segment as well us mild dilatation the bile to carcinoid tumor qf‘the a historic of left to metastntic curcirloid tumor.

of strictures,

dilatation,

filling

struction in each patient

Fig. I. ERCP (a) und MRC (bi show u normul intruheputic und extraheputic biliury tree. A normul pancrrutic duct is also visible on MRC.

522

CT scan, ERCP/PTC), by the Mayo

surgical

of ob-

MR cholangiography in PSC

The study population consisted of 74 patients. MRC could not be performed in a patient who became claustrophobic; hence, our data comprised 73 patients; 33 male, 40 female, mean age 56 years (range 19-94). The final diagnosis was determined by ERCP in 66 patients (90%); during surgery in five patients (7%); and by clinical and biochemical follow-up in two patients (3%).

_ _ . Statistical analysis Sensitivity, specificity, positive and negative predictive values, and diagnostic accuracy of MRC in the evaluation of the biliary tree were calculated using a “two-by-two” table. Findings by invasive cholangiography (ERCP and/or PTC) were used as the gold standard of refereke-for the recognition of dilatation, obstruction, presence and location of stones, visualization of gallbladder and cystic duct, and diagnosis of PSC. The overall usefulness of MRC and invasive cholangiography was determined correlating the final diagnosis on MRC and invasive cholangiography findings. The extra information provided by MRC (diagnostic) and ERCP/PTC (diagnostic and therapeutic)

Fig. 4. Typical changes of primary sclerosing cholangitis on MRC and ERCP. A4RC (a) shows a long stricture involving the common hepatic duct. However, on ERCP (b) this stricture is in fact tight and band-like. Since A4RC visualizes the bile ducts in their physiologic state, the bile duct segment distal to this high-grade stricture is collaused and - appears as a part of the stricture on MRC. Conversely. because distension of the distal bile duct segment occurs during$lling at ERCP, a better anatomic delineate of this biliary stricture is accomplished on ERCP.

523

P. Angulo et al.

Fig. 5. Cholangiocarcinoma developed in the setting of long-standing primary sclerosing cholangitis. ERCP (a) and MRC (b and c (magnified view)] show an abrupt narrowing of the CBD about 4 cm proximal to the ampulla, creating an appearance like a concentric lesion (arrows). The diagnosis of cholangiocarcinoma was established histologically.

Table 1 shows the final diagnosis in the patient population. Forty-two patients (58%) had benign biliary disease, including 23 patients (32%) with PSC; nine patients (12%) had malignant biliary disease; and 22 patients (30%) had a normal biliary tree. All patients with PSC had intrahepatic and extrahepatic involvement as shown on invasive cholangiography. Diagnostic quality images were obtained in 70173 (96%) of invasive cholangiography (68 ERCPs, 2 PTCs) and in 73/73 (100%) of MRC procedures. ERCP was unsuccessful in five patients, including three patients due to failure to cannulate the CBD (one patient had advanced gastric cancer with extensive involvement of the antrum and pylorus, another patient had cancer of the head of the pancreas with extensive gastric invasion, and the other patient had cancer of the head of the pancreas with extensive involvement of the distal CBD); in a patient with a duodenal diverticulum, and in a patient with PSC, ERCP was unsuccessful and had to be converted to PTC. A good-quality visualization of the biliary and pancreatic ducts was

TABLE Overall

3 comparison

of MR cholangiography

and invasive cholangiography

Sensitivity

MRC ERCPlPTC

524

obtained by MRC in these five patients in whom ERCP could not be performed. Using ERCP or PTC findings as the reference standard (70 patients), the sensitivity, specificity, predictive values, and diagnostic accuracy of MRC were determined and shown in Table 2. MRC had a diagnostic accuracy greater than 90% in the correct identification of normal bile ducts (Fig. l), biliary dilatation or obstruction (Fig. 2) and biliary stones as well as in the correct diagnosis of PSC (Fig. 3-5). MRC had a low sensitivity in the detection of biliary stones (50%) in the whole group of patients with a specificity of 98% and diagnostic accuracy of 97%. However, the sensitivity of MRC in the detection of bile duct stones was greater in patients without PSC compared to PSC patients (80% vs 20%). Stones not identified by MRC (false negatives) in five patients, including four patients with PSC, were located in the intrahepatic biliary system. Conversely, four out of five patients with stones correctly identified by MRC (true positives) were located in the common bile duct. MRC correctly de-

45150 (90%) 50/52 (96%)

Specificity

21123 (91%) 21/21 (100%)

using the final diagnosis Positive predictive

value

45147 (96(X>) 50/50 ( 100%)

as the standard

Negative predictive

value

2106 (81%) 21123 (91%)

of reference

Diagnostic accuracy 66173 (90%) 71/73 (97%)

(n=73)

MR cholangiography TABLE

Discussion

4

Diagnostic and/or patients (530/o)*

in PSC

therapeutic

interventions

during

ERCP

in 36/68

No. of interventions Patients with PSC (16/22) (73%) Balloon dilatation Brushings/biopsies Stones extraction Sphincterotomy Stent removal Patients without PSC (20/46) (43%) Stent placement Balloon dilatation Stent removal Biopsies Sphincterotomy

8 6 5 5 2 10 7 5 2 2

* The number of patients with additional interventions during ERCP was significantly greater in the group of patients with PSC (16/22, 73%) compared to patients without PSC (20/46, 43%), p=O.O2.

tected the location of biliary obstruction in 36 out of 37 patients (sensitivity of 97%). The gallbladder was identified on MRC in 26 out of 40 patients (65%) without history of cholecystectomy, and in 21 out of these 40 patients (52%) on ERCP The gallbladder identified on ERCP in these 21 patients was also visualized on MRC (sensitivity 100%). Using the final diagnosis (73 patients), Table 3 gives the overall comparison of MRC and invasive cholangiography. As shown, MRC had an excellent diagnostic accuracy in the detection of biliary disease that was similar to invasive cholangiography. Additional diagnostic and/or therapeutic interventions were performed in 36 out of 68 patients (53%) during the ERCP procedure, including 20 out of 46 patients (43%) without PSC and in 16 out of 22 patients (73%) with PSC (p=O.O2) as shown in Table 4. Balloon dilatation, stent placement or removal and histologic sampling were the most common interventions. The MRC procedure gave additional diagnostic information (3 1 findings) in 17 patients (23%) such as presence of a cirrhotic liver (11 patients), splenomegaly (seven patients), portal hypertensive circulation (four patients), hepatic/renal cysts (five patients), ascites (two patients), cavernous hemangioma (one patient), and portal lymph nodes (one patient) as well as a complete visualization of the intrahepatic biliary ducts above high-grade strictures not seen during ERCP in three patients with PSC. Six out of 68 patients (8.8%) who underwent ERCP developed complications such as abdominal pain (three patients) requiring hospitalization for at least 24 h, pancreatitis (two patients), and perforation of the CBD (one patient).

The invasive nature of ERCP or PTC along with their inherent technical difficulties and risks has prompted investigators to evaluate the usefulness of non-invasive imaging techniques such as MRC in patients with suspected biliary disease. In this study, we have confirmed and extended the results of previous reports demonstrating the excellent accuracy of MRC in the correct diagnosis of normal bile ducts (9) as well as bile duct abnormalities such as dilatation (7,9,10), obstruction (9,11-13), and bile duct stones (10,14). Our results also show that MRC is a useful imaging modality for the diagnosis of PSC. In patients with PSC, ERCP is frequently limited by the multiplicity of the bile duct strictures and the difficulty of opacifying the ducts proximal to high-grade stenoses. PTC is necessary in some patients for a better visualization of this more peripheral biliary tree. MRC, however, can provide a complete mapping of the entire biliary tree in patients with PSC, and it is of particular help in those patients in whom ERCP or PTC is unsuccessful or provides only partial visualization of the bile ducts due to complete bile duct obstruction. Thus, MRC can provide helpful clinical information before an invasive procedure such as ERCP or PTC. Similarly, since essentially all patients who have had an ERCP or PTC will have this invasive procedure preceded by an US or CT scan, MRC can give most of the diagnostic information with only one study. Hence, MRC might make other non-invasive imaging studies such as US and CT scan unnecessary. This should have financial benefits as well as saving the patient time, although this needs to be evaluated prospectively Likewise, in patients with an established diagnosis of PSC, MRC may become useful for disease follow up and monitoring. Although the sensitivity of MRC in detecting bile duct stones in patients without PSC was high (80%) and greater than that reported with ultrasonography or CT scan (15-17), our results show that MRC may have a low sensitivity in the correct detection of bile duct stones in the presence of PSC compared to patients without PSC. This may be expected particularly in patients in whom extensive obliterative changes of the biliary tree may mask the presence of bile duct stones, especially those with an intrahepatic location. Furthermore, MRC may be less sensitive in the correct identification of small (‘3 mm) bile duct stones, or small biliary strictures (9,11,12,18). MRC had an accuracy of 93% (65/70) and a sensitivity of 83% (19123) in the correct diagnosis of PSC in our study. Two of the four false negative cases had very mild changes of PSC, causing no dilatation on ERCP; MRC was interpreted as normal in these two cases. Although biliary stric-

525

P. Angulo et al.

tures and dilatation were identified on MRC in the other two false negative cases, the lack of clinical information at the time of MRC interpretation did not allow us to establish a definite diagnosis on MRC. Similarly, the single false positive PSC case was a patient who developed biliary strictures post-liver transplantation; the biliary strictures were correctly identified on MRC, but the diagnosis suggested on this study was PSC in the absence of clinical information about previous transplantation. Because MRC visualizes the bile ducts in their physiologic state, the bile duct segment immediately distal to a high grade stricture may be collapsed and appears as a part of the stricture, suggesting the image of a long stricture on MRC (Fig. 4). Hence, in some patients, MRC does not provide a detailed analysis of stricture morphology necessary in the differential diagnosis of a benign stricture, or in the setting of cholangiocarcinoma complicating PSC. Conversely, some distension of the ducts is inevitable at ERCP/PTC during filling. In patients with PSC who have extensive involvement of the biliary tree, such distension may provide a better anatomic delineation of some segments of the biliary tract, which can lead to a better identification of bile duct stones and better anatomic definition of bile duct strictures. Furthermore, patients with PSC often require diagnostic and/or therapeutic interventions during the course of the disease, such as brush cytology, biopsies, stone extraction, balloon dilatation, and, in some cases, stenting which cannot be accomplished during the MRC procedure. Thus, in patients with biliary disease including PSC, MRC can replace diagnostic cholangiography in most cases, whereas invasive cholangiography (ERCP, PTC) should be reserved for therapeutic indications. Development of cholangiocarcinoma represents the most lethal biliary complication in patients with PSC (4). Unfortunately, most patients with this biliary malignancy are diagnosed when the tumor is too advanced to allow successful resection, as happened in the single case of PSC plus cholangiocarcinoma in the current study. Further studies are needed to determine the role of MRC in the diagnosis of cholangiocarcinoma as well as the role of this imaging technique in improving prognosis. In summary, MRC and invasive cholangiography have excellent diagnostic accuracy in the presence of biliary disease. Although additional information may be provided by MRC, such as the status of the liver and spleen, presence of ascites, or adenopathy, and MRC would be of obvious benefit when ERCP is unsuccessful or inadequate, invasive cholangiography is the procedure of choice when therapeutic intervention 526

is anticipated. In contrast, because of its non-invasive nature, MRC has clear-cut advantages over invasive cholangiography in patients with suspected biliary disease, in particular PSC, when a complete visualization of the biliary tract is necessary for diagnostic purposes and/or for planning therapeutic interventions. The future of MRC, however, will depend not only on its diagnostic accuracy, but also on the availability of the technique and the cost at which the information is obtained. Although the MRC standard cost for Medicare is less than of ERCP, cost-effectiveness and outcome analysis studies are necessary to address these issues.

Acknowledgement Presented American (AASLD),

in part during the Annual Meeting of the Association for the Study of Liver Diseases Chicago IL, November 1998.

References of 1. Bilbao MK, Dotter CT, Lee TG, Katon RM. Complications endoscopic retragrade cholangiopancreatography (ERCP). A study of 10.000 cases. Gastroenterology 1976; 70: 314-20. DB. Riccobono C. Aaron B. Compli2. Zimmon DS. Falkenstein cations of endoscopic retrograde cholangiography. Analysis of 300 consecutive cases. Gastroenterology 1975; 69: 303-9. 3. Okuda K, Tanikawa K, Emurd T. Nonsurgical percutaneous transhepatic cholangiography: diagnostic signiiicance in medical problems of the liver. Am J Dig Dis 1974: 19: 21 -36. 4. Angulo P, Lindor KD. Primary sclerosing cholangitis. Hepatology 1999: 30: 325~ 32. 5. Reinhold C, Bret PM. Current status of MR cholangiopancreatography. AJR 1996: 166: 1285-95. 6. Soto JA. Barish MA, Yucel EK, Siegenberg D, Ferrucci JT, Chuttani R. Magnetic resonance cholangiography: comparison with endoscopic retrograde cholangiopancreatography. Gastroenterology 1996: 110: 589-m97. 7. Regan F. Smith D. Khazan R. Bohlman M. Schultze-Haakh H, Campion J, et al. MR cholangiography in biliary obstruction using half-Fourier acquisition. J Comput Assist Tomogr 1996: 20: 621?32. 8. Hintze RE. Alder A, Veltzke W. Abou-Rebyeh H, Hammerstingl R.Volgl T. et al. Clinical significance of magnetic resonance cholangiopancreatography (MRCP) compared to endocopic retrograde cholangiopancreatography (ERCP). Endoscopy 1997; 29: 182 7. 9. Mendler MH. Bouillet P, Sautereau D. Chaumerliac P, Cessot F, Sidaner AL, et al. Value of MR cholangiography in the diagnosis of obstructive disease of the biliary tree: a study of 5X cases. Am J Gastroenterol 1998: 93: 2482-90. 10. Holrknccht N. Gauger J, Sackmann M, Thoeni RF Schurig J. Ho11 J. et al. Breath-hold MR cholangiography with snapshot techniques: prospective comparison with endoscopic retrograde cholangiographp Radiology 1998; 206: 657 64. 11. Cuilbaud L. Bret PM. Reinhold C. Atri M. Barkun AN. Bile duct obstruction and choledocholitiasis: diagnosis with MR cholangiography. Radiology 1995: 197: 109%15. 12. lshinzaki T, Wakayamu T. Okada Y. Kobayashi T. Magnetic resonance cholangiography for evaluation of obstructive jaundice. Am J Gastroenterol 1993; 88: 2072-7. 13. Hall-Cragg MA. Allen CM. Owens CM, Theis BA, Donald JJ, Paley M, et al. MR cholangiography: clinical evaluation in 40 cases. Radiology 1993: 189: 423-7. 14. Bret PM, Reinhold C. Magnetic resonance cholangiography: a field with attraction. Gastroenterolocgy 1995: 108: 1949%50.

MR cholangiography 15. Pasanen P Partanen K, Pikkarainen P Alhava E, Pirinen A, Janatuinen E. Ultrasonography, CT, and ERCP in the diagnosis of choledochal stones. Acta Radio1 1992; 33: 5336. 16. Stott MA, Farrands PA, Guyer PB, Dewbury KC, Browning JJ, Sutton R. Ultrasound of the common bile duct in patients undergoing cholescystectomy. J Chn Ultrasound 1991; 19: 73-6. 17. Regan F, Fradin J, Khasan R, Bohlman M, Magnuson T. Chole-

in PSC

docholitiasis: evaluation with MR cholangiography. AJR 1996; 161: 144-5. 18. Low RN, Sigeti JS, Francis IR, Weinman D, Bower B, Shimakawa A, et al. Evaluation of malignant biliary obstruction: efficacy of fast multiplanar spoiled gradient-recalled MR imaging vs spinecho MR imaging. CT, and cholangiography. AJR 1994; 162: 315-23.

527