Copy of pediatric MRCP

Heavily T2-weighted fast spin-echo FSE and single-shot FSE MR imaging sequences with long echo times are used to image the biliary and pancreatic ducts.. Factors that affect image qualit

1951 EDUCATION EXHIBIT

Govind B Chavhan, MD, DNB • Paul S Babyn, MD • David Manson,

MD • Logi Vidarsson, PhD

High-quality magnetic resonance (MR) cholangiopancreatographic im-ages are difficult to obtain in children due to the small caliber of the pediatric bile ducts and to motion artifacts However, there has been ongoing improvement in image quality, thanks to better coil technology, increased speed of acquisition, refinement in respiratory compensation techniques, and newer sequences Heavily T2-weighted fast spin-echo (FSE) and single-shot FSE MR imaging sequences with long echo times are used to image the biliary and pancreatic ducts Secretin has been shown to improve the visualization of the pancreatic duct and pan-creaticobiliary junction Factors that affect image quality in pediatric

MR cholangiopancreatography include sedation, negative oral contrast material, radiofrequency coil selection, respiratory compensation tech-niques, echo time, echo train length, section-slab thickness, planes of scanning, field of view, and number of signals acquired However, giv-ing proper attention to these factors and tailorgiv-ing the study to the body size of the patient (which varies considerably) can lead to high-quality diagnostic MR cholangiopancreatographic images Use of MR chol-angiopancreatography in children is limited by the need for sedation

or anesthesia, high cost, limited availability, and long scanning times Nonetheless, this modality can be a viable alternative to endoscopic ret-rograde cholangiopancreatography (ERCP) in the evaluation of various entities such as choledochal cyst, recurrent pancreatitis, primary scleros-ing cholangitis, and a transplanted liver, and may obviate ERCP.

©RSNA, 2008 • radiographics.rsnajnls.org

Pediatric MR Chol- angiopancreatogra-phy: Principles, Tech-nique, and Clinical

ONlINE-ONly

CME

See www.rsna

.org/education

/rg_cme.html

lEARNING

OBJECTIVES

After reading this

article and taking

the test, the reader

will be able to:

Discuss the

■

principles,

imag-ing sequences, and

clinical applications

of MR

cholangio-pancreatography in

children.

List various

fac-■

tors affecting the

quality of MR

cholangiopancreato-graphic images.

Describe potential

■

challenges to

ob-taining good-quality

MR

cholangiopan-creatographic

im-ages in children and

ways to overcome

them.

Abbreviations: CBD = common bile duct, ERCP = endoscopic retrograde cholangiopancreatography, ETL = echo train length, FSE = fast

spin-echo, MIP = maximum-intensity-projection, PBJ = pancreaticobiliary junction, PSC = primary sclerosing cholangitis, RF = radiofrequency, SNR = signal-to-noise ratio, TSE = turbo spin-echo, 2D = two-dimensional, 3D = three-dimensional

RadioGraphics 2008; 28:1951–1962 • Published online 10.1148/rg.287085031 • Content Codes:

1 From the Department of Diagnostic Imaging, The Hospital For Sick Children and University of Toronto, 555 University Ave, Toronto, ON, Canada M5G 1X8 Presented as an education exhibit at the 2007 RSNA Annual Meeting Received February 19, 2008; revision requested March

18 and received March 31; accepted April 9 All authors have no financial relationships to disclose Address correspondence to G.B.C (e-mail:

drgovindchavhan@yahoo.com).

© RSNA, 2008

distribution to your colleagues or clients, use the RadioGraphics Reprints form at the end of this article.

See last page

TEACHING

POINTS

Magnetic resonance (MR)

cholangiopancreatog-raphy can be an effective, noninvasive imaging

tool for the evaluation of pancreaticobiliary

dis-ease in children (1–5) Its diagnostic accuracy is

reported to exceed 90% compared with direct

cholangiopancreatography in conditions such

as choledocholithiasis, absent gallbladder, and

pancreas divisum (2) and to approach 100% in

conditions such as choledochal cyst (4,6)

How-ever, it is challenging to perform a good-quality

MR cholangiopancreatographic examination in

children, especially very young children

Pedi-atric MR cholangiopancreatography is limited

by small-caliber ducts, poor signal, and patient

motion, which creates artifacts Nonetheless, it is

possible to visualize ducts as small as 1 mm in

di-ameter, thanks to improvements in coil

technol-ogy, increased speed of acquisition, refinements

in respiratory compensation techniques that

reduce motion artifacts, and newer sequences

(7,8) In a large series of 85 pediatric MR

chol-angiographic studies performed in 78 patients

(mean age, 10.3 years), excellent image quality

was seen in 85% of cases (1)

With the continuous improvement in image quality, more and more studies on MR

cholan-giopancreatography in children are appearing in

the literature (1–6) In this article, we review the

basic principles of MR cholangiopancreatography

and the types of imaging sequences used In

ad-dition, we discuss and illustrate various factors

affecting MR cholangiopancreatographic image

quality, with emphasis on potential challenges

and ways to overcome them We also discuss the

current clinical applications of pediatric MR

cholangiopancreatography

Principles of MR Cholangiopancreatography

At MR cholangiopancreatography, the bile

within the biliary tree is imaged with heavily

T2-weighted sequences The sequences are heavily

T2 weighted with use of long echo times in the

range of 300–1000 msec, such that only tissues

or fluid with prolonged transverse relaxation time

(T2) retain signal These tissues and fluid are

seen as hyperintense structures The background

soft tissues with shorter T2 do not retain

signifi-cant signal long enough in a sequence with

pro-Types of Sequences Used Fast Spin-Echo

(Turbo Spin-Echo) Sequence

With the fast spin-echo (FSE) (turbo spin-echo [TSE]) sequence, a 90° radiofrequency (RF) pulse is followed by a train of 180° RF pulses

The number of 180° pulses is called the echo train length (ETL) or turbo factor The speed of acquisition increases with an increase in ETL

For the purposes of MR cholangiopancreatogra-phy, FSE imaging is performed with a long ETL, repetition time, and echo time (>250 msec) (7)

A long ETL is well suited for the type of long-echo-time acquisition required for MR chol-angiopancreatography FSE imaging can be per-formed with a breath hold or with free breathing with use of respiratory gating or a navigator tech-nique It can be performed as a two-dimensional (2D) sequence with a slab thickness of 2–7 cm or

as a three-dimensional (3D) sequence with thin-ner sections The resultant data can then be re-constructed into cholangiographic images with a maximum-intensity-projection (MIP) technique

Fast recovery FSE (FRFSE; GE Medical Systems, Waukesha, Wis), RESTORE (Siemens Medical Solutions, Forchheim, Germany), and driven equilibrium FSE (DRIVE; Philips Medi-cal Systems, Best, the Netherlands) sequences are modified FSE sequences in which a 90° RF pulse

is used at the end of the ETL to get the magneti-zation immediately in longitudinal plane, thereby reducing repetition time

Single-Shot FSE (TSE) Sequence

Single-shot FSE (SSFSE, GE Medical Systems), half-Fourier single-shot TSE (HASTE, Siemens), and single-shot TSE (SSTSE, Philips) sequences are FSE sequences in which just more than one-half of k-space is filled, thus reducing scanning time significantly All of the required k-space lines (phase-encoding steps) are filled in a single repetition time; hence the name “single-shot.”

Acquisition times are in seconds, with reasonably good spatial resolution (9) SSFSE imaging has a higher signal-to-noise ratio (SNR) than does FSE imaging as a result of less motion artifact pro-ducing noise (7) Limitations of SSFSE imaging

longed echo time and are, therefore, suppressed

Blood vessels are not seen, since flowing blood

does not produce any signal on these images

Teaching Point

Teaching

Point

include image blur induced by long ETLs, flow

artifacts, and problems with saturation of

adja-cent sections (9)

Echo times typically range between 300 and

1000 msec These sequences can be performed

with a breath hold or with respiratory triggering,

and good-quality images can be obtained, even

with quiet breathing

Other Sequences

Fast gradient-echo sequences such as balanced

fully refocused steady-state sequences (true fast

imaging with steady-state precession, fast

imag-ing employimag-ing steady-state acquisition, balanced

fast field echo) show the biliary tree well, with

excellent SNR and good spatial resolution These

sequences can be performed with a breath hold

or with quiet breathing and may show ducts

reasonably well when other sequences fail to do

so because of motion artifacts One limitation,

however, is the fact that blood vessels are seen as

bright structures, which may make it difficult to

differentiate them from bile ducts

Contrast Material–enhanced Functional

Evaluation.—Three-dimensional fast

gradi-ent-echo T1-weighted sequences such as fast

low-angle shot (FLASH, Siemens) and spoiled

gradient-echo (SPGR, GE Medical Systems)

se-quences can be performed after the intravenous

injection of a contrast agent such as mangafodipir

trisodium (Teslascan; Nycomed Amersham,

Princeton, NJ) that is excreted via the biliary

system This technique offers a few additional

advantages, including assessment of biliary

func-tion, less problematic background suppression of

ascites and bowel fluid, and identification of

po-tential biliary leaks (9) Mangafodipir trisodium

has been used in neonates to study biliary atresia

(10)

Secretin MR Cholangiopancreatography.—

Secretin is a polypeptide hormone secreted by

duodenal mucosa in response to acid stimulation

It increases the pancreatic secretion of water and

bicarbonate, improves the tone of the sphincter

of Oddi, and increases the amount of fluid in the

duodenum (1) Thus, the normal response to

se-cretin administration is increased fluid signal in

the pancreatic duct and subsequent fluid

excre-tion into the duodenum Secretin is administered

intravenously in a dose of 1 unit per kilogram

of body weight slowly over 1 minute (11,12) It has been used in children in doses of 0.2 µg per kilogram of body weight (maximum dose, 16 µg) injected intravenously slowly over 1 minute (1) Images obtained with fast T2-weighted se-quences, usually constituting a coronal slab along the pancreatic duct, are acquired every 30 sec-onds for 10 minutes Normal response includes increased signal and an increase in the diameter

of the pancreatic duct to as much as 3 mm that peaks 3–5 minutes after secretin injection, fol-lowed by a progressive return to the baseline diameter within 10 minutes (11,12) Persistent dilatation of the pancreatic duct after 10 minutes

is considered an abnormal response and can

be seen in stricture, sphincter of Oddi dysfunc-tion, or papillary stenosis (13) Secretin has been shown to improve visualization of the pancreatic duct and its junction with the common bile duct (CBD) (1,2,12) It also improves the sensitivity

of MR cholangiopancreatography in diagnosing early-onset idiopathic chronic pancreatitis (11)

A major disadvantage of secretin MR cholangio-pancreatography is the high cost of secretin

Problems in Obtaining

MR Cholangiopancreatograms

The smaller duct caliber in children makes it dif-ficult to visualize nondilated ducts at MR cholan-giopancreatography, especially when evaluating intrahepatic duct caliber Lack of signal, along with reduced spatial resolution, can make the visualization of ducts more difficult Image qual-ity is often poor because of motion and breath-ing artifacts Even with respiratory triggerbreath-ing, image quality can be poor, since it is difficult to achieve respiratory cycles with a regular rhythm and adequate amplitude in infants and small children The need for sedation or general anes-thesia is another major drawback of pediatric MR cholangiopancreatography

Factors Affecting

MR Cholangiopancre- atographic Image Quality Sedation

We sedate most neonates, infants, and young children Children under 6 months of age are sedated with oral chloral hydrate (50–100 mg

and the phase-encoding steps that can be used Both of these factors improve image quality but

at the cost of increased scanning time

Respiratory triggering can be performed in two ways In the first method, respiratory tracings are obtained by tying a bellows around the chest, and images are acquired in a designated phase

of respiration with each respiratory cycle (Fig 1) The second method of respiratory triggering is the navigator technique With this technique, the position of the diaphragm is detected with a navi-gator pulse, and signal acquisition is prospectively

or retrospectively gated to the most stable portion

of the respiratory cycle, typically end expiration (Fig 1b) (9)

Acquisition time with respiratory gating is typ-ically 3–7 minutes but may last for more than 10 minutes if breathing is not regular Section mis-registration due to irregular breathing can also occur with some respiratory gating techniques

per kilogram of body weight) Children between

6 months and 6 years of age are usually sedated

with the intravenous injection of midazolam

(Versed; Hoffman–La Roche, Basel,

Switzer-land) (0.05 mg per kilogram of body weight) or

pentobarbital (Nembutal; Ovation

Pharmaceu-ticals, Deerfield, Ill) (5 mg per kilogram of body

weight) Most children over 6–8 years of age can

cooperate sufficiently without sedation

Negative Oral Contrast Material

Negative oral contrast material can be used in

nonsedated children to reduce the high signal of

gastric secretion and intestinal fluid Its use has

been shown to improve image quality without

causing any significant adverse reactions (1,6,11)

Ferumoxsil and ferric ammonium citrate are two

commonly used negative oral contrast agents

Ferumoxsil consists of superparamagnetic iron

oxide particles and is administered as 150–300

mL of oral suspension (1,11) Ferric ammonium

citrate, a T2-negative contrast agent, is mixed

with water (1 g/10 mL) and administered orally

(1 mL per kilogram of body weight) (6) Oral

negative contrast agent is given just before the

examination, when the patient is placed on the

table

RF Coil Selection

The smallest coil that fits the pediatric patient

be-ing imaged should be selected as the receiver coil

Smaller coils permit use of smaller fields of view

and offer better resolution The noise detected

with the coil increases with coil size Infants and

neonates can be imaged in a quadrature knee

coil, a head coil, or even flexible surface coils

Older children are usually imaged with torso

phased-array coils The body coil is used only for

transmission; its use as a receiver coil should be

avoided if possible because signal quality is poor

Respiratory

Compensation Techniques

MR cholangiopancreatographic sequences

performed with a breath hold provide the

best-quality images However, not all patients can hold

their breath, especially younger children, infants,

and neonates Moreover, breath-hold technique

limits the number of signals that can be acquired

Figure 1 Respiratory triggering methods (a)

Draw-ing illustrates how a respiratory tracDraw-ing is obtained by tying a bellows over the chest Images are acquired in

a designated phase of respiration with each respiratory

cycle (b) Image illustrates the navigator technique,

in which the position of the diaphragm is detected with

a navigator pulse, and signal acquisition is prospec-tively or retrospecprospec-tively gated to the most stable por-tion of the respiratory cycle Box and arrow indicate the position of the navigator where the navigator pulse will hit.

an echo time of 300–600 msec provides a good balance between signal in the ducts and tissue suppression

Echo Train Length

ETL is the number of 180° RF pulses sent after

a 90° pulse in a single repetition time in FSE sequences As ETL increases, scanning time is reduced; however, image blurring also increases Moderate ETL in the range of 16–20 should be used, especially with respiratory-gated FSE T2-weighted sequences that are not constrained by

Echo Time

Echo time is an important parameter in MR

cholangiopancreatographic sequences As echo

time increases, background tissue suppression

increases, improving visualization of the bile

ducts Higher echo times above 1000 msec

re-duce overall signal and may complicate the

vi-sualization of small ducts containing only small

amounts of bile in the pediatric population (Fig

2) Hence, the echo time used in these patients

should be moderate The use of a wide range of

echo times (180–1000 msec) for pediatric MR

cholangiopancreatography has been reported in

the literature (1,6,11,14–19) In our experience,

Figure 2 Effect of changes in echo time Coronal FSE T2-weighted MR images of the biliary tree obtained with echo times of 180 (a), 400 (b), 600 (c), and 1100 (d) msec (a, b, and d obtained in the same patient) show how

background tissue suppression increases as echo time increases Arrow indicates the CBD.

Slabs with a 2–7-cm thickness can be acquired

in straight coronal or axial planes or in a radiat-ing fashion in coronal oblique planes Coronal oblique sections are acquired at the porta hepatis and radiating from a point anterior to the portal vein bifurcation (Fig 4a) (7) Straight coronal and initial left posterior oblique images more clearly depict the anteriorly located common hepatic duct, left hepatic duct, and proximal pancreatic duct, whereas the more posteriorly located CBD, right hepatic ducts, and distal pancreatic duct including the ampulla are seen on left posterior oblique images obtained at a steeper angle (7) Most of the pancreatic duct may be visualized in

an axial oblique plane (Fig 4b) Segmental liver transplants are better evaluated in sagittal oblique planes, since the “neo” porta hepatis is oriented

in a more anteroposterior direction (19)

Other Parameters

The field of view should be adjusted to the size

of the child A field of view that is too small, together with a large matrix, may make images grainy and reduces SNR A 256 × 256 matrix is

Figure 3 Effect of changes in section-slab

thick-ness Coronal SSFSE (GE Medical Systems)

im-ages obtained with an echo time of 400 msec and

section-slab thicknesses of 5 (a), 20 (b), and 50 (c)

mm show how tissue overlap increases as section

thickness increases Arrow indicates the CBD

GB = gallbladder.

breath-hold time With single-shot sequences,

ETL usually equals the number of k-space lines

to be filled

Section-Slab Thickness

Sections obtained with 3D FSE sequences should

be as thin as possible without any gap In fact, an

overlap of sections is preferable

With 2D FSE and single-shot sequences, thin

sections in the range of 3–5 mm as well as slabs

ranging from 2 to 7 cm in thickness are acquired

Slab thickness should be tailored to the size of

the child such that a maximum volume of the

biliary tree is obtained without significant

over-lap by other structures (Fig 3) For infants and

neonates, a slab thickness less than 2 cm is

usu-ally sufficient Thin sections are used for smaller

filling defects in ducts and provide better spatial

resolution However, thinner sections also have a

lower SNR

Planes of Scanning

Three-dimensional FSE sequences are usually

performed in the coronal plane MIP reformatted

images can then be obtained in any plane

Two-dimensional FSE and SSFSE

thin-sec-tion images are typically acquired in the axial and

coronal planes For evaluation of tiny calculi or

ductal filling defects, the axial plane is preferred

improvement is proper coil selection Either a phased-array or a surface coil that fits the child well should be used for signal reception Poor signal in children can be compensated for to some extent by increasing the number of signals averaged to between four and eight However,

this needs to be balanced with the ETL, since

in-creasing the number of signals averaged increases scanning time Proper slab thickness in 2D FSE and SSFSE imaging is the next step toward ob-taining optimal-quality images Smaller children such as neonates and infants do not need a slab thickness of more than 2 cm Thin-section (3–5-mm) images in the axial and coronal planes, as well as thick slabs in radiating coronal planes, should be acquired A 3D FSE sequence should

be performed whenever breathing is regular

It provides better SNR and spatial resolution than does 2D FSE imaging (Fig 5) Moreover, anomalies can be better understood by rotating 3D images Because neonates and infants usually have irregular breathing with varying respiratory amplitude, 3D FSE imaging with respiratory trig-gering may not be possible in many of these pa-tients (16) SSFSE imaging with varying section thickness can be useful in this situation, provid-ing adequate image quality even with free shallow breathing Alternatively, balanced steady-state free precession, with its excellent SNR, insensitiv-ity to motion, and speed of acquisition, can be

typically used, with a pixel size ranging between

1 and 1.5 mm2 An increased number of signals

averaged (about four to six) can be used in

pedi-atric MR imaging to improve the signal

Optimization of Pediatric

MR Cholangiopancreatography

Pediatric MR cholangiopancreatography needs

to be tailored to the different body sizes intrinsic

to children of various ages For visualization of smaller-caliber ducts, spatial resolution and SNR need to be improved The first step toward such

Figure 4 Planes of scanning (a) Axial FRFSE (GE Medical Systems) image shows coronal oblique planes

radiat-ing from a point anterior to the portal vein bifurcation (b) Axial oblique SSFSE image shows nearly the entire

nor-mal pancreatic duct (snor-mall arrows) and a prominent CBD (large arrow).

Figure 5 MIP image from 3D FSE imaging data

shows a normal CBD (long arrow) and part of the

pancreatic duct (short arrow) Note the improved

spa-tial resolution and SNR of the image.

Teaching Point

Teaching Point

atography is not routinely used in biliary atresia due to its high cost and the need for sedation

Choledochal Cyst

Choledochal cysts represent common congeni-tal abnormalities that may be associated with complications such as infection, obstruction, and development of malignancy Therefore, complete cyst excision without compromising the pancre-atic duct and common pancrepancre-aticobiliary channel should be performed as soon as the diagnosis is made to reduce the risk of malignancy later in life (4,23) The diagnosis of choledochal cyst is usu-ally made with US However, information about the type of cyst, the length of the involved duct, the presence and location of protein plugs or cal-culi, the pancreaticobiliary junction (PBJ), and the length of the common channel is required for preoperative planning ERCP has traditionally been used to obtain this information However,

MR cholangiopancreatography has been shown

to be 100% accurate in the evaluation of chole-dochal cyst (Fig 6) (2,4,6)

Evaluation of the PBJ

Abnormal union of the PBJ with a long common channel has been implicated as the cause of cho-ledochal cyst formation and as one of the causes

of pancreatitis in children (5,6) It has been hy-pothesized that abnormal union of the PBJ out-side the duodenal wall and the sphincter of Oddi allows the reflux of pancreatic exocrine enzymes into the biliary duct, causing cholangitis and in-creased pressure leading to dilatation of the bile ducts Conversely, reflux of bile into the pancre-atic duct can cause pancreatitis (6,24) Reports vary with respect to the detection rate of abnor-mal union of the PBJ at MR cholangiopancre-atography, which ranges from 40% to 83% (6) It

is difficult to visualize abnormal union of the PBJ

in children under 2 years of age and in those with

a large choledochal cyst, especially with cystic dilatation overlapping the PBJ (6,25) Secretin

MR cholangiopancreatography has been shown

to improve the visualization of abnormal union

of the PBJ (1,2,4) A common pancreaticobiliary channel more than 15 mm in length is consid-ered abnormal in adults (24) To our knowledge, however, there are no data on the normal length

of the common channel in children (Fig 7a) The mean length of the common channel increases with age (6) A length of 5 mm was used as the upper limit of “normal” for the common channel

in pediatric studies by Fitoz et al (4) and Kim et

al (25) (Fig 7b)

used as a problem-solving sequence A 3-T

im-ager provides superior-quality MR

cholangiopan-creatographic images because it has double the

SNR of a 1.5-T imager

Clinical Applications

In spite of its capacity to provide anatomic and

pathologic details of the biliary tree, MR

cholan-giopancreatography is not widely used because of

the ready availability of other modalities such as

ultrasonography (US) and scintigraphy, as well as

the frequent need for sedation, high cost, limited

availability, and time required for scanning With

recent improvements in image quality and

resolu-tion, however, MR cholangiopancreatography is

increasingly being used and has become a viable

alternative to endoscopic retrograde

cholangio-pancreatography (ERCP) and percutaneous

transhepatic cholangiography for diagnostic

pur-poses As mentioned earlier, in a series of 85

pe-diatric MR cholangiographic studies performed

in 78 patients, excellent image quality was seen in

85% of cases (1) Concordance with ERCP

find-ings was seen in 81% of cases, with additional

findings that could not have been visualized with

ERCP alone seen in 35 of 85 studies (1) These

findings included hepatosplenomegaly, hepatic

tumors, pancreatic masses, varices, adrenal

hem-orrhage, multicystic dysplastic kidney, and

omen-tal and mesenteric metastases MR

cholangiopan-creatography should be considered before ERCP

and percutaneous transhepatic cholangiography

In the following sections, we discuss a variety of

applications of MR cholangiopancreatography in

children

Biliary Atresia

The extrahepatic bile ducts are almost always

visualized at MR cholangiopancreatography

Biliary atresia can be ruled out if the entire

extrahepatic bile duct is seen (16) In a study

of 26 infants (mean age, 2 months), MR

chol-angiography was shown to be 82% accurate,

90% sensitive, and 77% specific for depicting

extrahepatic biliary atresia (20) However, MR

cholangiopancreatography faces strong

competi-tion from (a) scintigraphy, which has excellent

sensitivity (96%–97%) in differentiating biliary

atresia from neonatal hepatitis (21), and (b) US,

in which constant improvements are being made

A constellation of various signs at US have been

found to be highly accurate (98%) in the

detec-tion of biliary atresia (22) MR

cholangiopancre-Teaching Point

abnormal union of the PBJ, and pancreas di-visum MR cholangiopancreatography is useful for noninvasively identifying these structural anomalies and ruling them out as a cause of pancreatitis (5,24,27) It can be performed in

Recurrent Pancreatitis

Common causes of pancreatitis in children

include trauma, structural pancreaticobiliary

anomalies, systemic diseases, infection, and

drugs; however, recurrent pancreatitis may be

idiopathic (26) Common structural anomalies

causing pancreatitis include choledochal cyst,

Figure 6 Type IV choledochal cyst in a 13-year-old girl MIP image from 3D FSE imaging data (a) and

coronal fast imaging employing steady-state acquisition image (b) show a large cystic structure representing a

grossly dilated CBD Moderate dilatation of the intrahepatic ducts is also seen, along with downward

displace-ment of the pancreatic duct (arrowheads in a).

Figure 7 (a) Normal union of the PBJ in a 9-year-old child Coronal thick-slab SSFSE image shows normal

union of the PBJ (arrowhead) The common hepatic duct is mildly dilated dd = duodenum, PD = pancreatic duct

(b) Abnormal union of the PBJ in an 11-year-old boy with a choledochal cyst Coronal short inversion time

inver-sion-recovery image shows a dilated CBD (curved arrow) containing multiple calculi A long common pancreatico-biliary channel (arrowhead) inferior to the junction of the CBD and the pancreatic duct (straight arrow) is also seen.

in such cases Cholangiographic findings of PSC include irregularity of the bile duct wall; areas of mild dilatation, with intermittent strictures giving the bile ducts a “beaded” appearance; formation

of sacculations and pseudodiverticuli; and nonvi-sualization of the intrahepatic ducts (Fig 9) (29)

Other Conditions Affecting the Liver and Pancreas

MR cholangiopancreatography can be useful in the evaluation of pathologic conditions such as

the acute stage of pancreatitis—unlike ERCP,

which is contraindicated in the acute stage MR

cholangiopancreatography is highly sensitive and

specific for pancreas divisum (Fig 8) (24) It is

difficult to visualize pancreatic ducts and

anoma-lies such as pancreas divisum and (as mentioned

earlier) anomalous union of the PBJ in children

less than 2 years of age (6) It has been reported

that the pancreatic duct is visible at MR

cholan-giopancreatography in 45%–60% of cases (4)

Again, secretin has been shown to improve

visu-alization of the pancreatic duct and its junction

with the CBD (1,2,12) Secretin also improves

the sensitivity of MR cholangiopancreatography

in diagnosing early-onset idiopathic chronic

pan-creatitis (11) and pancreas divisum (12)

Primary Sclerosing Cholangitis

Primary sclerosing cholangitis (PSC) is an

id-iopathic condition characterized by obliterative

fibrosis and inflammation of the bile ducts It is

most commonly seen in patients with

inflamma-tory bowel disease such as ulcerative colitis PSC

is being recognized with increasing frequency

in children ERCP is the standard of reference

for establishing the diagnosis, since pathologic

changes are nonspecific In a study by Ferrara et

al (28), MR cholangiopancreatography was found

to have a specificity and positive predictive value

of 100% and an accuracy of 85% The authors

of that study concluded that positive MR

chol-angiopancreatographic findings in a child with

clinical suspicion for PSC are very likely to be

correct; hence, ERCP should not be performed

Figure 8 Pancreas divisum in a 9-year-old girl Coronal SSFSE images show that the CBD is opening at the ma-jor papilla (arrow in a) and the pancreatic duct draining the body is opening at the minor papilla (arrowhead in b)

The pancreatic duct normally joins the CBD and opens at the major papilla dd = duodenum.

Figure 9 PSC in a 14-year-old boy Coronal MIP

image from 3D FSE T2-weighted imaging data shows dilated extra- and intrahepatic ducts with multiple strictures, outpouchings, and peripheral cystic dilata-tions The cystic duct (long arrow) is also dilated and irregular and joins the bile duct (short arrows) at a point inferior to the normal point of juncture The pancreatic duct (arrowheads) is normal.