Magnetic Resonance Cholangiopancreatography

However, ERCP is highly operator dependent, has sig-nificant morbidity and mortality, and operators cannot can-nulate the common bile duct CBD and pancreatic duct in up to 9% of examinat

Ahmet Mesrur Halefoglu, MD

Magnetic resonance cholangiopancreatography (MRCP)

is a noninvasive imaging technique for the evaluation

of pancreatico-biliary disorders It uses magnetic resonance

imaging to visualize fluid in the biliary and pancreatic ducts

as high signal intensity on T2-weighted sequences and

pro-vides improved spatial resolution and permits imaging of the

entire pancreatico-biliary tract during a single breath-hold It

is being used with increasing frequency as a noninvasive

alternative to diagnostic endoscopic retrograde

cholangio-pancreatography (ERCP) and, in most institutions, has

be-come the initial imaging tool for the pancreatico-biliary

sys-tem, with ERCP reserved for only therapeutic indications

The literature indicates that MRCP is equivalent in

diag-nostic accuracy to ERCP across numerous pancreatico-biliary

pathologies and therefore can reliably be used as the first-line

investigation MRCP is noninvasive, less operator dependent,

does not require anesthesia or contrast material, and uses no

radiation It is only a diagnostic procedure and therapeutic

intervention cannot be performed as part of this procedure,

whereas ERCP is a diagnostic as well as a therapeutic

proce-dure However, ERCP is highly operator dependent, has

sig-nificant morbidity and mortality, and operators cannot

can-nulate the common bile duct (CBD) and pancreatic duct in

up to 9% of examinations.1

MRCP avoids the potential morbidity and mortality

asso-ciated with ERCP MRCP is an appropriate noninvasive tool

in suspected pancreato-biliary pathology especially when no

or low likelihood of therapeutic intervention is anticipated

MRCP is particularly useful where ERCP is difficult,

danger-ous, or impossible (eg, previous gastroenteric anastomosis or

gastrojejunostomy) It is also an important option for patients

with failed ERCPs MRCP and ERCP have different

contrain-dications allowing them to be used as complementary

tech-niques.2In this article, we provide an overview of the MRCP

technique and clinical applications in a variety of diseases

Technical Aspects

The MRCP technique relies on the use of heavily T2-weighted imaging sequences, which display stationary fluid (ie, bile and pancreatic secretions) as areas of high signal intensity MRCP was initially performed with gradient-echo sequences These were generally slow and gave poor quality images When the fast spin-echo pulse sequence became available, it replaced the gradient-echo MRCP technique The advent of fast sequences has led to dramatically shortening the imaging time and has provided breath-hold techniques viable Single shot fast spin-echo (SSFSE) (or half-Fourier turbo spin-echo) technique is a variant of fast spin-echo technique and is currently the sequence of choice for MRCP.3 These techniques allow cholangiographic images to be obtained in a very short breath-hold Rapid imaging avoids motion arti-facts (eg, related to bowel peristalsis, respiration, and volun-tary motion) and allows noncooperative patients to be eval-uated

Optimal MRCPs are required with a high-field scanner, a torso phased-array coil, and fast (breath-hold) sequences MRCP is usually performed initially with a single-shot pro-jection technique in which thick-slab (40-70 mm) axial and coronal images of the upper abdomen are obtained to localize the extrahepatic bile duct Next, a multisection technique involving the acquisition of multiple thin-slab source images (3 to 5 mm) in the coronal oblique plane along the longitu-dinal axis of the bile duct is performed Three-dimensional (3D) images can be generated from these source images with

a maximum-intensity projection (MIP) algorithm Although the thick collimation and 3D MIP images more closely resem-ble conventional cholangiograms and are familiar to many clinicians, spatial resolution is degraded because of volume-averaging effects Diagnostic decisions should be made on the basis of the source images, although these cholangio-gram-like MIP images are very helpful in providing an over-view of ductal anatomy Thin source images are shown to be more sensitive than the MIPs in detecting small calculi.4 The breath-hold technique is superior to non-breath-hold techniques in that it eliminates artifacts arising from respira-tory motion In addition, the use of phased-array surface coils has resulted in improved image quality by increasing signal-to-noise ratios Because of improvements in image quality, MRCP is capable of showing ducts as small as 1 mm.3,5An

Sisli Etfal Training and Research Hospital, Department of Radiology,

Istan-bul, Turkey.

Address reprint requests to Ahmet Mesrur Halefoglu, MD, Sisli Etfal

Train-ing and Research Hospital, Department of Radiology, 34360, Sisli,

Istan-bul, Turkey E-mail: halefoglu@hotmail.com

282 0037-198X/08/$-see front matter © 2008 Elsevier Inc All rights reserved.

doi:10.1053/j.ro.2008.06.004

additional advantage of the SSFSE or half-Fourier turbo

spin-echo technique is the ability to reduce suspectibility effects

from surgical clips, metallic biliary and vascular stents,

bili-ary drainage catheters, and spinal fixation rods The

reduc-tion of suspectibility effects is important because MRCP is

often performed in patients who have multiple clips

second-ary to cholecystectomy, bilisecond-ary-enteric anastomosis, or liver

transplantation

A routine MRCP protocol for a 1.5-Tesla scanner in our

institution is shown inTable 1 Gadolinium-enhanced

im-ages of the liver and pancreas may also be added to the

protocol in the case of suspected tumor This uses a 3D

spoiled gradient echo T1 sequence before and after

gadolin-ium administration, with images acquired 20 and 70 seconds

after the start of the bolus gadolinium injection

Patient Preparation

MRCP can be performed in all patients apart from those with

specific internal ferro-magnetic foreign bodies or

claustro-phobia Patients should be fasted approximately 4 to 6 hours

before examination to provide gallbladder filling and gastric

emptying IV contrasts or antispasmodics are not used No

exogenous contrast material is needed to demonstrate the

pancreatico-biliary system Similarly, glucagon use is not

rec-ommended to obviate peristalsis due to rapid enough pulse

sequences Administration of a negative contrast agent can be

helpful by providing reduction of the signal intensity from

overlapping fluid-filled structures such as the stomach and

duodenum

Normal Anatomy

The gallbladder and CBD are visualized in up to 98% of

patients6(Fig 1) Visualization of the biliary tree is variable

distal to the right and left hepatic ducts.6In addition, MRCP

is 95% accurate in differentiation of normal from dilated

ducts.7MRCP using SSFSE technique allows visualization of

the entire normal-caliber pancreatic duct in the head and

body in 97% of cases and in the tail in 83% of cases5(Fig 1)

Complete visualization of a dilated pancreatic duct is possible

in 100% of cases Demonstration of pancreatic side branches

varies from 19% in the head to 5% in the tail.1

Clinical Applications

Choledocholithiasis

MRCP is very helpful in the diagnosis of choledocholithiasis because CBD stones appear as low-signal-intensity foci within the high-signal-intensity bile (Fig 2) Calculi as small

as 2 mm in diameter can be visualized.8,9Small calculi may not cause secondary dilation of the ducts9and are best seen

on the axial images.9It is crucial to scrutinize the thin, source images because the sensitivity for detection of small stones decreases with an increase in section thickness owing to vol-ume averaging of high-signal-intensity bile surrounding the stone

Shanmugam and coworkers10assessed the predictive value

of MRCP in the diagnosis of choledocholithiasis MRCP find-ings were compared with ERCP or operative findfind-ings Of the

221 patients, the MRCP showed a sensitivity of 97.98% and specificity of 84.4% The authors stated that MRCP is highly sensitive and specific for choledocholithiasis and avoids the

Table 1 MRCP Protocol (optimized for 1.5-Tesla scanner)

1 Localizer: Coronal oblique SSFSE T2-weighted 2D sequence (TR: 1200 msec; TE: 140 msec; Bandwidth: 31.25; Section thickness: 70 mm; Intersection gap: 1.5 cm; FOV: 44 mm; Matrix: 320 ⴛ 192; Slice number: 20; Scanning time: 24 seconds)

2 Axial images: Axial fast spin-echo, respiratory triggered T2-weighted 2D thin-slabe fat-suppressed sequence (TR: 6600 msec; TE: 90 msec; Echo train length: 13; Bandwidth: 41.67; Section thickness: 3 mm; Intersection gap: 1-1.5 mm; NEX: 3; FOV: 44 mm; Matrix: 256 ⴛ 224; Slice number: 24; Scanning time: 5 minutes).

3 Coronal oblique images along the expected angle of the common and pancreatic ducts: Coronal SSFSE T2-weighted 2D thick-slab fat-suppressed breath-hold sequence (TR: 2715 msec; TE: 1360 msec; Bandwidth: 31.25; Section thickness: 50 mm; Intersection gap: 0; FOV: 34 mm; Matrix: 388 ⴛ 288; Slice number: 6; scanning time: 16 seconds)

Figure 1 Coronal MIP image Gallbladder (GB), cystic duct (CD),

common bile duct (CBD), and pancreatic wirsung duct (PD) are clearly seen in this normal patient.

need for invasive imaging in most patients with suspected

choledocholithiasis

Griffin and coworkers11prospectively assessed the

accu-racy of MRCP in diagnosing bile duct stones as an alternative

to ERCP in 115 patients with suspected CBD stones awaiting

laparoscopic cholecystectomy MRCP showed a sensitivity of

84%, specificity of 96%, positive-predictive value of 91%,

negative-predictive value of 93%, and diagnostic accuracy of

92% when compared with ERCP as the gold standard The

authors concluded that MRCP can be reliably used as the

first-line investigation for choledocholithiasis

Barish and colleagues12state that MRCP can visualize the

normal or dilated CBD in 96 to 100% of patients Stones

appear as areas of signal void within the high-signal-intensity

bile on MRCP The authors note the sensitivity of MRCP for

detecting choledocholithiasis has been reported to be

be-tween 71 and 100%

Benign Biliary Strictures

More than 80% of bile duct strictures occur after an injury to

the extrahepatic bile ducts during a cholecystectomy, with a

minority attributable to other benign causes such as

infec-tion, pancreatitis, stone passage, trauma, primary sclerosing

cholangitis, ischemia, chemotherapy, and acquired

immuno-deficiency syndrome MRCP and ERCP have complementary

roles in the diagnosis of biliary strictures Although normal

and dilated CBDs are consistently demonstrated on MRCP,

early strictures that have not yet caused biliary dilation are

relatively difficult to demonstrate on MRCP Once the biliary

tree is dilated, MRCP performs well.13

In many cases, it is very difficult to distinguish between

benign and malignant etiologies Benign strictures tend to be

longer with more gentle sloping shoulders than malignant strictures and do not have associated masses (Fig 3)

Intrahepatic Duct Disease

The role of MRCP in the evaluation of intrahepatic duct dis-ease (eg, primary sclerosing cholangitis (PSC), acquired

im-Figure 2 (A) Axial thin slab source image shows multiple hypointense calculi in the gallbladder and CBD (B) Coronal

MIP image clearly demonstrates CBD stone as a hypointense round lesion located in the distal region.

Figure 3 Coronal MIP image CBD distal segment stenosis (arrow) is

seen secondary to infection in a patient.

munodeficiency syndrome cholangiopathy) is increasing.

Preliminary data suggest that MRCP may be used to establish

the diagnosis of PSC14and obviate diagnostic ERCP

The performance of MRCP in PSC depends on the severity

of disease When there are focal strictures with intervening

dilated segments, the diagnosis is readily made on MRCP

(Fig 4), but in earlier phases of the disease, early stenoses

may be missed and short strictures may be overestimated

because the downstream duct is collapsed Nevertheless,

good correlation between ERCP and MRCP images in the

diagnosis of PSC has been shown in a published study.15In

another recently published case-control study, Moff and

co-workers16reviewed 36 patients with PSC to determine the

diagnostic accuracy of both MRCP and ERCP The authors,

based on their study, stated that MRCP could be a useful

screening test for PSC, but ERCP should remain as the

con-firmatory test, given its higher specificity Currently the role

of MRCP is confined to the follow-up of advanced cases

and/or the development of complications

Cholangiocarcinoma

Cholangiocarcinoma arises from the bile duct epithelium and

may occur anywhere along the intrahepatic or extrahepatic

bile ducts, from the liver to the ampulla of vater MRCP plays

an important role in the assessment of perihilar

cholangio-carcinoma and in many institutions it has replaced ERCP and

percutaneous transhepatic cholangiography for the

preoper-ative staging of the tumor.17 In cholangiocarcinoma cases,

MRCP can accurately depict the presence and level of

ob-struction17 and has been shown to be more effective than

ERCP in delineating the anatomic extent of the cancerous

infiltration18 (Fig 5) Romagnuolo and coworkers19

per-formed a literature search to estimate the overall sensitivity

and specificity of MRCP in suspected biliary obstruction

They concluded that MRCP seems to be highly accurate for

diagnosing the presence of obstruction, but it is less accurate

at differentiating malignant from benign causes of

obstruc-tion

The combination of parenchymal and vascular

informa-tion obtained from the T1-weighted, T2-weighted, and

gad-olinium-enhanced images, and bile duct information ob-tained from the MRCP images, can be used to accurately stage cholangiocarcinoma MRCP images alone are not adequate to identify the cause of biliary obstruction due to the fact that they solely provide luminal information, and gadolinium-enhanced images are necessary for complete evaluation of the biliary obstructions

Inflammatory Changes

Pancreatitis is the most common benign disease involving the pancreas and is classified as acute or chronic on the basis of clinical, morphological, and histologic criteria

Acute Pancreatitis

In patients presenting with acute pancreatitis, the detection

of gallstones and the state of the pancreatico-biliary tree are of major importance MRCP provides the opportunity to ac-quire similar diagnostic information to ERCP in this regard without risk The pancreatic duct may be normal in mild cases of acute pancreatitis Occasionally, the enlarged and edematous pancreas can cause compression of the pancreatic duct In these cases, the pancreatic duct is either not visual-ized or presents a smooth and symmetric narrowing MRCP provides information relating to ductal dilation, ductal dis-ruption, leakage, peripancreatic fluid collections, and intra-ductal lesions predisposing to pseudocyst formation.20 MRCP can easily detect pancreatic pseudocysts, their shape, number, and size, providing valuable information for the surgeon (Fig 6)

Figure 5 Coronal MIP image Klatskin tumor is visible, causing

ex-tensive dilation of the main and intrahepatic biliary ducts.

Figure 4 Axial thin slab source image, multiple strictures, and

dila-tations of the intrahepatic biliary ducts leading to a beaded

appear-ance typical of primary sclerosing cholangitis in a patient.

Chronic Pancreatitis

Chronic pancreatitis represents irreversible exocrine damage

to the pancreas and irreversible morphologic changes in the

pancreas and pancreatic duct These changes include dilation

of the main pancreatic duct and its side branches and contour

irregularities In severe pancreatitis, side branches have a

“chain of lake” appearance Additional pancreatic ductal

changes include stricture formation and intraductal calculi

These calculi are seen as low-signal-intensity filling defects

surrounded by high-signal-intensity pancreatic fluid

(menis-cus sign) Stones as small as 2 mm can be detected by MRCP.5

In advanced chronic pancreatitis cases, the pancreatic duct

dilation is more pronounced and can be accompanied by

CBD dilation producing “double duct sign” as in the case of

pancreatic head carcinoma (Fig 7)

Sica and coworkers21compared MRCP with ERCP in 30

patients with chronic pancreatitis and in 9 with acute

pan-creatitis MRCP sensitivity was found to be 91% They

con-cluded that in patients with pancreatitis, MRCP provides

di-agnostic information similar to that with ERCP and thus

could be used similarly to guide patient treatment MRCP is

also helpful in all patients with technically failed ERCP

ex-aminations

In another study performed by Soto and coworkers,22

they found the sensitivity of MRCP for ductal dilation as

87-100%, for ductal narrowing as 75%, and for ductal

calculi as 100% They concluded that MRCP can

accu-rately demonstrate pancreatic duct abnormalities in

chronic pancreatitis

Congenital Abnormalities

MRCP can be used to demonstrate a variety of congenital anomalies of the pancreatico-biliary tract MRCP has been shown to be 98% accurate in diagnosis of aberrant hepatic ducts and 95% accurate in diagnosis of cystic duct variants.23

A potential use of MRCP is in the evaluation of bile duct anatomy before cholecystectomy By demonstrating aberrant anatomy before surgery, the risk of bile duct injury can be reduced Anatomic variants with a high potential for injury include an aberrant right hepatic duct with insertion into the common hepatic duct, or a cystic duct inserting medially on the CBD.24

Pancreas divisium is the most common pancreatic congen-ital anomaly and results from failure of fusion of the dorsal and ventral pancreatic ducts Its prevalence is around 10% The larger dorsal duct drains the tail, body, and superior part

of the head of the pancreas and passes anterior to the distal CBD to end at the minor papilla The smaller ventral duct drains the inferior head and uncinate process and joins with the CBD to exit via the major papilla (Fig 8) Although this variant may be detected incidentally in asymptomatic pa-tients, pancreas divisium occurs more frequently in patients who present with acute idiopathic pancreatitis than in the general population.25MRCP has a sensitivity of up to 100%

in its detection.26 Annular pancreas is seen in 1 of every 20,000 autopsies and is characterized by pancreatic tissue completely or in-completely surrounding the duodenum, most commonly the descending duodenum Definitive diagnosis relies on ERCP demonstration of the annular pancreatic duct MRCP now allows the diagnosis of this anomaly noninvasively.27

Figure 7 Coronal MIP image Both CBD and pancreatic duct dilation

leading to double duct sign are demonstrated in this chronic pan-creatitis patient.

Figure 6 Coronal thin slab source image A large pseudocyst

forma-tion located at the head of the pancreas (arrow) seen in an acute

pancreatitis patient.

Malignant Neoplasms

The majority of pancreatic malignant tumors are ductal

ade-nocarcinoma and between 60 and 70% of these

adenocarci-nomas are located in the head of the pancreas MRCP is useful

in the evaluation of pancreatic adenocarcinoma and

intra-ductal papillary mucinous tumors The typical MRCP ap-pearance of pancreatic head carcinoma is represented by sud-den obstruction at the level of the head of the pancreas with

a double duct sign, due to biliary and pancreatic duct dila-tion, and evidence of mass effect (Fig 9).This sign is highly suggestive but not specific to malignancy and occurs in 77%

of cases.28The morphology of the obstruction can be helpful

in the differential diagnosis between pancreatitis and neo-plastic lesion, although not pathognomonic In general, the obstruction secondary to pancreatic cancer presents with a

“mouse tail” pattern or with sudden reduction of the caliber

of the bile duct In the case of pancreatitis the biliary duct stenosis has a tapered aspect Regarding pancreatic duct, in neoplastic lesions it is usually homogenously dilated, whereas in chronic pancreatitis an irregular dilation with a beaded appearance can be seen

Intraductal papillary mucinous tumors are slow-growing tumors and produce large amounts of mucin They originate from the main pancreatic duct or side branches epithelium These tumors are seen as cystic side branches dilation or grape-like lesions with a communicating channel with the main pancreatic duct (Fig 10) MRCP can be regarded supe-rior to ERCP in the diagnosis, because mucin often impedes contrast filling of these ducts

In ampullary carcinoma cases, together with the CBD ob-struction, high-grade obstruction with abrupt termination accompanying dilation of the pancreatic duct is usually prominent29(Fig 11)

Postsurgical Biliary Tract Alterations

MRCP plays a critical role in evaluating the surgically altered biliary tract ERCP is often difficult or impossible to perform

Figure 8 Coronal thick slab image A dorsal pancreatic duct (PD)

passing anterior to the common bile duct (CBD) and draining into

the minor papilla through the small duct of Santorini (SD) is

dem-onstrated A smaller ventral pancreatic duct (VD) and CBD draining

into the major papilla can also be seen.

Figure 9 (A) Axial thin slab source image, dilation of the CBD, and pancreatic duct leading to a double duct sign due to

pancreatic head carcinoma is seen (B) Coronal thick slab image A huge pancreatic head mass causing marked dilation

of the CBD and pancreatic duct is demonstrated on the same patient.

in patients with biliary-enteric anastomoses, including

cho-ledochojejunostomy, hepaticojejunostomy, and Billroth type

2 gastrectomy MRCP is now the technique of choice in this

situation, with a sensitivity of 100% in demonstrating the

biliary-enteric anastomoses.5

Pavone and coworkers30 used MRCP to examine 24

pa-tients with biliary-enteric anastomoses and noted a sensitivity

of 100% in detecting anastomotic strictures and of 90% in

detecting biliary tract stones proximal to the anastomoses

MRCP is also 100% sensitive in demonstrating the

cho-ledochojejunal anastomosis after a whipple procedure5

(Fig 12)

Pancreatic Trauma

Traumatic injuries to the pancreatic duct may be related to penetrating or blunt trauma The pancreatic duct may also be injured during surgery, particularly splenectomy Barkin and coworkers31reported a sensitivity and specificity of 100% for ERCP for the detection of pancreatic duct disruption In some instances, MRCP may show the duct disruption as well as associated fluid collections Houben and coworkers32 con-ducted a retrospective study including 15 children who had pancreatic trauma Both computed tomography (CT) and MRCP were performed MRCP was performed in seven chil-dren with four who were also subjected to ERCP for compar-ison MRCP correctly predicted the nature of the duct injury

It was also useful in correctly predicting absence of duct injury in one patient whose CT findings were suggestive of a pancreatic duct injury The authors concluded that a mini-mally invasive approach avoiding the need for open surgery

is possible but relies on accurate definition of the degree of pancreatic trauma using a combination of contrast-enhanced

CT and MRCP imaging, predicting the need for ERCP Soto and coworkers33 in their series including seven trauma patients accurately demonstrated the status of pan-creatic duct and the site of duct injury in all patients by MRCP

Advantages and Limitations

The contraindications to MRCP are the presence of specific ferro-magnetic objects within the body, such as pacemakers

or aneurysms clips Claustrophobia is the most common cause of unsuccessful examination

Figure 10 Coronal MIP image A cystic dilation communicating with

a channel to main pancreatic duct representative of intraductal

pap-illary mucinous tumor is seen.

Figure 11 Coronal MIP image Both CBD and pancreatic duct

dila-tion is seen in this patient who has peri-ampullary carcinoma.

Figure 12 Coronal thick slab image CBD is seen draining into the

jejunum following choledochojejunostomy and remnant pancreatic duct is seen draining into the jejunum, after pancreaticojejunos-tomy in a patient who underwent whipple operation due to pancre-atic head adenocarcinoma.

Reduced spatial resolution of MRCP in comparison with

ERCP can cause difficulty in some situations With MRCP, it

may be challenging to detect the early changes in sclerosing

cholangitis, with a tendency to either overlook or,

con-versely, overestimate the length of short strictures Similarly,

in the assessment of pancreatitis, more subtle side branch

changes are sometimes not resolved by MRCP in comparison

with ERCP

The major advantage of MRCP is the lack of invasiveness

The other advantages are operator independence, easily

im-plementation of pulses sequences, and application for

pa-tients with altered anatomy where ERCP cannot be

per-formed

Conclusion

MRCP is a noninvasive important tool in the diagnosis of

pancreatico-biliary diseases and a promising alternative to

ERCP In many institutions, MRCP is replacing diagnostic

ERCP as the modality of choice for pancreatico-biliary

imag-ing However, it should be remembered that, unlike ERCP,

MRCP does not allow the opportunity to simultaneously

per-form therapeutic intervention Knowledge of the advantages

and disadvantages of each technique is needed to determine

the appropriate workup of patients with pancreatico-biliary

disease

References

1 Ueno E, Takada Y, Yoshida I, et al: Pancreatic diseases: evaluation with

MR cholangiopancreatography Pancreas 16:418-426, 1998

2 Halefoglu AM: Magnetic resonance cholangiopancreatography: a useful

tool in the evaluation of pancreatic and biliary disorders World J

Gas-troenterol 13:2529-2534, 2007

3 Irie H, Hoda H, Tajima T, et al: Optimal MR cholangiopancreatographic

sequence and its clinical application Radiology 206:379-387, 1998

4 Yamashita Y, Abe Y, Tang Y, et al: In vitro and clinical studies of image

acquisition in breath-hold MR cholangiopancreatography: single-shot

projection technique versus multislice technique AJR Am J Roentgenol

168:1449-1454, 1997

5 Fulcher AS, Turner MA, Capps GW, et al: Half-Fourier RARE MRCP in

300 objects Radiology 207:21-32, 1998

6 Vitellas KM, Keogan MT, Spritzer CE, et al: MR cholangiography of bile

and pancreatic duct abnormalities with emphasis on the single-shot fast

spin-echo technique Radiographics 20:939-957, 2000

7 Soto JA, Barish MA, Yucel EK, et al: Magnetic resonance

cholangiogra-phy: comparison to endoscopic retrograde cholangiopancreatography.

Gastroenterology 110:589-597, 1996

8 Fulcher AS, Turner MA: MR pancreatography: a useful tool for

evalu-ating pancreatic disorders Radiographics 19:5-24, 1999

9 Becker CD, Grossholz M, Becker M, et al: Choledocholithiasis and bile

duct stenosis: diagnostic accuracy of MR cholangiopancreatography.

Radiology 205:523-530, 1997

10 Shanmugam V, Beattie GC, Yule SR, et al: Is magnetic resonance

cholangiopancreatography the new gold standard in biliary imaging ?

Br J Radiol 78:888-893, 2005

11 Griffin N, Wastle ML, Dunn WK: Magnetic resonance

cholangiopan-creatography versus endoscopic retrograde cholangiopancholangiopan-creatography

in the diagnosis of choledocholithiasis Eur J Gastroenterol Hepatol 15:809-813, 2003

12 Barish MA, Yucel EK, Ferrucci JT: Magnetic resonance cholangiopan-creatography N Engl J Med 341:258-264, 1999

13 Schwartz LH, Coakley FV, Sun Y, et al: Neoplastic pancreaticobiliary duct obstruction: evaluation with breath-hold MR cholangiopancre-atography AJR Am J Roentgenol 170:1491-1495, 1998

14 Ernst O, Asselah T, Talbodec N, et al: MR cholangiopancreatography: a promising new tool for diagnosing primary sclerosing cholangitis AJR

Am J Roentgenol 168:1115-1116, 1997

15 Ernst O, Asselah T, Sergent G, et al: MR cholangiography in primary sclerosing cholangitis AJR Am J Roentgenol 171:1027-1030, 1998

16 Moff SL, Kamel IR, Eustace J, et al: Diagnosis of primary sclerosing cholangitis: a blinded comparative study using magnetic resonance cholangiography and endoscopic retrograde cholangiography Gastroi-ntest Endosc 64:219-223, 2006

17 Lee SS, Kim MH, Lee SK, et al: MR cholangiography versus cholangios-copy for evaluation of longitudinal extension of hilar cholangiocarci-noma Gastrointest Endosc 56:25-32, 2002

18 Yeh TS, Jan YY, Tseng JH, et al: Malignant perihilar biliary obstruc-tion: magnetic resonance cholangiopancreatographic findings Am J Gastroenterol 95:432-440, 2000

19 Romagnuolo J, Bardou M, Rahme E, et al: Magnetic resonance cholan-giopancreatography: a meta-analysis of test performance in suspected biliary disease Ann Intern Med 139:547-557, 2003

20 Takehara Y: MR pancreatography Semin Ultrasound CT MR

20:324-339, 1999

21 Sica GT, Braver J, Cooney MJ, et al: Comparison of endoscopic retro-grade cholangiopancreatography with MR cholangiopancreatography

in patients with pancreatitis Radiology 210:605-610, 1999

22 Soto JA, Barish MA, Yucel EK, et al: Pancreatic duct: MR cholangiopan-creatography with a three dimensional fast spin-echo technique Radi-ology 196:459-464, 1995

23 Taourel P, Bret PM, Reinhold C, et al: Anatomic variants of the biliary tree: diagnosis with MR cholangiopancreatography Radiology 199: 521-527, 1996

24 Martin RF, Rossi RL: Bile duct injuries: spectrum, mechanisms of in-jury, and their prevention Surg Clin North Am 74:781-803, 1994

25 Bernard JP, Sahel J, Giovanni M, et al: Pancreas divisium is a probable cause of acute pancreatitis: a report of 137 cases Pancreas 5:248-254, 1990

26 Bret PM, Reinhold C, Taorel P, et al: Pancreas divisium: evaluation with

MR cholangiopancreatography Radiology 199:99-103, 1996

27 Hidaka T, Hirohashi S, Uchida H, et al: Annular pancreas diagnosed by single-shot MR cholangiopancreatography Magn Reson Imaging 16: 441-444, 1998

28 Barish M, Soto J, Ferrucci J: Magnetic resonance pancreatography En-doscopy 29:487-495, 1997

29 Semelka R, Kelekis N, John G, et al: Ampullary carcinoma: demonstra-tion by current MR techniques J Magn Reson Med 7:153-156, 1997

30 Pavone P, Laghi A, Catalano C, et al: MR cholangiography in the exam-ination of patients with biliary-enteric anastomoses AJR Am J Roent-genol 169:807-811, 1997

31 Barkin JS, Ferstenberg RM, Panullo W, et al: Endoscopic retrograde cholangiopancreatography in pancreatic trauma Gastrointest Endosc 34:102-105, 1988

32 Houben CH, Ade-Ajayi N, Patel S, et al: Traumatic pancreatic duct injury in children: minimally invasive approach to management J Pe-diatr Surg 42:629-635, 2007

33 Soto JA, Alvarez O, Munera F, et al: Traumatic disruption of the pan-creatic duct: diagnosis with MR pancreatography AJR Am J Roentgenol 176:175-178, 2001