MR Pancreatography

Yasuo Takehara
Recent innovations in the field of ultrafast MR imaging have increased the robustness of MR cholangiopancreatography (MRCP). Its complete noninvasiveness and flexible physiological approaches in detecting pancreaticobiliary
pathologic conditions are gaining the acceptance of many clinicians. The procedure is also safer and more
comfortable both for physicians and patients compared with direct pancreatography or cholangiography. Because
of its cost effectiveness and safety, optimized MRCP technologies will gradually replace the diagnostic use of
endoscopic retrograde cholangiopancreatography (ERCP). It is also notable that MRCP techniques can be used to
obtain physiological/dynamic information that ERCP cannot provide. This article addresses recent advances in
MRCP from technological and clinical aspects, focusing on its unique features as a hydrographic technique, and also
refers to its limitations.
Copyright 1999 by W.B. Saunders Company

'ONINVASIVE IMAGING of the pancreaticobiliary tree traditionally has been performed

by ultrasonography (US) and CT. 1-3 With these
widely used techniques, the level of ductal obstruction as well as the cause of obstruction is usually
defined. However, neither modality provides a
projectional overview of the pancreaticobiliary
system, and both modalities have limitations in the
documentation of fine ductal anatomy and pathologic conditions. Moreover, CT generally requires
the administration of iodinated contrast media,
which is relatively expensive,4,5 nephrotoxic, and
may cause allergic reactions. 6,7 US is safe and
inexpensive, but it is largely operator dependent.
Therefore, when an abnormality is detected by US
or CT, endoscopic retrograde cholangiopancreatography (ERCP) is currently the next diagnostic
choice at most institutions.
For decades, ERCP was the only nonoperative
technique that provided a direct view of the biliary
and pancreatic ducts. ERCP is a safe procedure
when performed by an experienced operator. Skilled
operators can successfully cannulate 90% to 95%
of pancreatic ducts and 90% of biliary ducts. 8
However, the reported success rate does not include
patients who declined ERCP because of acute
pancreatitis, age or infirmity, allergy to the contrast
media or premedications, prior upper gastrointestinal surgery, or partial resection of the pancreas. If
such technical failures are taken into account, the
performance of direct methods may be more lim-

From the Department of Radiology, Hamamatsu University

School of Medicine, Hamamatsu, Japan.
Address reprint requests to Yasuo Takehara, MD, ScD,
Department of Radiology, Hamamatsu University School of
Medicine, 3600 Handa, Hamamatsu, 431-3192, Japan.
Copyright 1999 by W.B. Saunders Company
0887-2171/99/2005-000451 O.00/0
324

ited than previously reported. Asymptomatic amylase elevations occur in 30% to 40% of patients
after pancreatography and are rarely of clinical
significance. Although mostly selflimiting, ERCPinduced pancreatitis occurs in 3% to 5% of patients, 9,1resulting in the majority of lawsuits in this
field. In a nationwide survey of complications
conducted in Japan, the morbidity rate was 0.117%
and mortality rate was 0.0067%. 11 Aside from
complications, ERCP examination is accompanied
by hazards of ionizing radiation to both patients
and operators, lz
MR cholangiopancreatography (MRCP) is based
on the hydrographic effect of bile fluids and
pancreatic juice as well as suppression of surrounding structures. 13,14 The fundamental concept of
hydrographic technique is simple and straightforward. The images are based on heavily T2weighted sequences emphasizing nonftowing fluid
within the fluid-containing structures, whereas solid
tissues and flowing blood have a negligibly small
signal. Consequently, the integrity of fluid-containing structures is evaluated indirectly. MRCP can
depict pancreaticobiliary ducts affected by various
sorts of pathologic conditions. It is completely
noninvasive and does not require administration of
contrast. MRCP can reveal luminal narrowing,
obstructions, irregularities in the ductal system
with stenosis, dilatation, sacculation, and ectasia of
the pancreatic duct. The most pronounced feature
of MRCP compared with ERCP is that it can
delineate the ductal system upstream of complete
obstructions. Unlike direct cholangiography or pancreatography, this modality can show the entire
ductal system in 1 image. Its ability to reflect
physiological conditions of exocrine pancreatic
excretion is also one of the unique features of this
method. Also, MRCP is highly sensitive for detecting any cystic lesions. 15 Its inherent capability as a

Seminars in Ultrasound, CT, and MRI, Vo120, No 5 (October), 1999: pp 324-339

MR PANCREATOGRAPHY

hydrographic image allows detection of an increased number of cystic lesions, which are often
overlooked by other modalities.
In the last 5 years, MRCP has gained rapid
acceptance by endoscopists and surgeons because
of the familiar projectional image format. They
acquire images of the biliary and pancreatic structures in their native configuration, without the
pressure distension of the ducts associated with
injection of contrast agents.
IMAGING TECHNIQUES

Because MRCP is dependent on heavily TZweighted imaging with fast MR sequences, several
different techniques that enhance the main pancreatic duct (MPD) depiction have been implemented.
Previously, these techniques used gradient-echo
steady state free precession (SSFP), I6'17 and more
recently, fast spin-echo (FSE) and its variants for
raw data acquisitions. With various techniques,
data may be acquired during either suspended
respiration 18,19 or free breathing, 2-24 on 2-dimensional Fourier transform (2DFT) 19'21'24'z5or 3-dimensional Fourier transform (3DFT), z,26,27 and with
either body coils or phased array multicoils, zs
Recent innovations in echo-planar technologies,
such as fast receiver and high performance gradient
systems, have further accelerated the speed of data
acquisition and image quality of MRCE z9
FSE
The advent of FSE has shortened imaging time
for T2-weighted images significantly. 3,31 Unlike
conventional spin-echo, multiple echoes with successive nvpulses are acquired in FSE, but a separate
phase encoding gradient is given before each echo.
Each detected echo represents a different line
within a single k-space. As multiple lines are filled
for each repetition of the sequence, resultant scan
time is significantly shortened. Application of multiple w pulses in FSE also minimizes susceptibilityinduced signal loss. 32'33 Because of its robustness,
FSE and its variants are normally used in MRCP
for raw data acquisitions? 4-38
2DFT Multislice FSE Versus 3DFT FSE
Initially, MRCP with 2DFT FSE sometimes
showed a pseudostenosis in the main pancreatic
duct on the reconstructed images ~9 (Fig 1A). This
artifact occttrs partly because of postprocessing and
partly because of physiological depiction of the

325

MRCR For instance, maximum intensity projection

results in loss of signal between images because of
cross talk between slices, misregistration, and sharing of MPD signal with contiguous sections. Thus,
postprocessed projection images often show duct
discontinuity that does not actually exist. This
artifact becomes conspicuous as the sections become thicker. 3DFT acquisition can improve spatial resolution in the slice-selective direction, z,27,39
Therefore, the pseudostenosis may appear less
frequently on MRCP reconstructed from 3DFT
data acquisition, z 3DFT data acquisition can also
improve the signal-to-noise ratio (SNR) at the cost
of a lengthy acquisition time. 4 If performed with
breath-holding, it requires multiple, segmented
breath-holding periods that may tire patients, or if
done with respiratory gating, it further extends the
imaging time.
Postprocessing
To enhance the clarity of reconstructed images,
several postprocessing techniques are sometimes
required. Maximum intensity projection algorithm
(MIP) enables a 3D projection overview of the
biliary and pancreatic duct from any arbitrary
viewpoints. MR hydrography nonselectively visualizes all stationary water, irrespective of its origin.
With appropriate postprocessing, such as targeted
MIP or selective MIP reconstruction, overlaps of
such signals, other than pancreaticobiliary origin,
can be eliminated to some extent. More recently, a
new reconstruction algorithm reflecting depth information was introduced. 4~ Most recently, however,
the ductal anatomy was also delineated in a projectional fashion using a single thick slab method
without any postprocessing.
Single-Shot FSE
The snap shot or single shot technique that
collects all of the phase encoding steps after a
single excitation allows motion artifacts to be
avoided while SNRs remain sufficient.18,3 Blurring
is a possible drawback of echo train imaging such
as FSE sequence or single-shot FSE (SSFSE) using
a high number of spin echoes. The blur appears in
the phase-encoding direction, and degrades the
image. 4z However, this blurring effect is more
pronounced in tissues with shorter T2 relaxation
times, hence it is less problematic with MRCR
Therefore, with reduced motion artifacts and acqui-

326

sition times, these snapshot sequences appear to be

ideally suited for MRCE

Single Thick Slab Versus Multiple Thin

Slice Method
The single thick slab projection MRCP using
SSFSE is a robust technique in delineating nondilated pancreatic ducts 36,43,44 (Fig 1B). Unlike the
projection images postprocessed through MIP, this
method does not suffer from pseudostenosis or
discontinuity of the duct. The single thick slab
method alone, however, is not appropriate for
investigating the intraductal pathologic conditions
within the dilated ducts. Normally, lesions such as
calculi and papillary tumors within the duct are
detectable as filling defects on MRCP, but because
a significant number of overlying water signals are
included in 1 thick section using this method, these

YASUO TAKEHARA

filling defects may be obliterated by partial volume

averaging.
With the projection technique, contrast-to-noise
ratio (CNR) of the pancreaticobiliary system decreases as slice thickness increases. Dilatation and
occlusion of the pancreaticobiliary tree are equally
well shown by single thick slab and multiple slice
MIP images. 38 However, abnormalities in the periampullary region and anomalies in the pancreaticobiliary tree are seen more clearly on single thick
slab images than on MIP images. 38 Stones in the
common bile duct, gallbladder, or intrahepatic bile
ducts, however, are best seen on source images
acquired by the multislice technique. 38
The single thick slab method with long echo time
(TE) is useful in overviewing the pancreaticobiliary ducts, however, the depiction is limited if the
target structures contain relatively short T2 values
produced by proteinaceous fluid or thick bile.
Another limitation of the single thick slab method
is that it cannot demarcate the boundaries of
organs. Combined use of multiple thin slice method
using SSFSE can provide this information. We take
multiple thin slices with medium TE so that thick
bile or proteinaceous fluid shows bright signals.
For this purpose, the TE should not exceed 100
milliseconds. The multiple thin slice method should
be done without fat saturation because the boundaries of organs are defined by surrounding fat
signals. Likewise, ductal walls are to be demarcated so that pancreatic stones are more clearly
localized within the pancreatic duct.

Multiangle Oblique Slab Technique With Long TE

Fig 1. A 76-year-old woman with chronic pancreatitis. (A)

MRCP postprocessed with MIP using coronal source images
obtained by 2DFT FSE sequence. There are several segments
of discontinuity in the MPD (arrows), which may be interpreted as pathological stenoses. (B) MRCP obtained with a
single thick slab method shows perfect delineation of the
entire MPD. ERCP proved no stenosis in the MPD of this
patient.

To cover the pancreas with a relatively complex

configuration using a single slab, the thickness
should be more than 50 mm in the majority of the
cases (Fig 2A). On single slab method, as slab
thickness decreases, the SNR of the pancreaticobiliary system increases. Therefore, to avoid discontinuity of the thin pancreatic duct and to exploit
better SNR, a series of muhiangle oblique slab
acquisitions is recommended29 (Fig 2B). The merits of this method are twofold. First, it can avoid
overlap with other fluid-filled structures such as
renal or hepatic cysts or gastric juice between the
folds; second, it can improve the delineation of the
side branches. As slab thickness decreases, ductal
delineation of the side branches is significantly
improved. 38 This method cannot delineate all pancreatic ducts in a single imaging slab; however,

MR PANCREATOGRAPHY

Fig 2. Single slab versus multiangle oblique slab selection

prescribed on the axial image. (A) The shaded area shows the
prescribed slab covering the pancreas with a single thick slab
method. To include the entire length of the pancreatic duct,
the slab thickness cannot be reduced to a thickness under 50
mm. (B) The shaded areas indicate multiangle oblique slabs
prescribed to include the pancreas. The thinner slabs can
improve the delineation of the MPD and side branches,

observation of all imaging slabs enables thorough

investigation of the pancreatic ducts.

Secretin Administration
As previously discussed in the "2DFT versus
3DFT FSE" segment, MRCP was formerly postprocessed by MIP using source images. The source
images were acquired by the multiple thin slice
method with SSFSE or conventional FSE. When
comparing MIP-reconstructed MRCP with ERCR
MRCP can correctly diagnose the dilated segments
shown by ERCP in the majority of cases; however,
it has been shown to overestimate stenosed segments, or detect a false-positive stenosis 19(Fig 1A).
This was partly caused by artifacts created in the
process of MIP reconstruction, and, in some cases,
MRCP might have been reflecting the real physiological conditions of the MPD in situ. The conditions include decreased output of pancreatic juice

327

in chronic pancreatitis or undilated MPD between

meals. In such conditions, one cannot differentiate
the pathological stenosis from physiologically undilated MPD. ERCP, on the other hand, can successfully inflate the unaffected segments using a pressurized injection of contrast media unless there is
complete obstruction downstream. The ductal segments unaffected by chronic pancreatitis are considered to be soft and flexible, thus dilatable by
secretin. The segments affected by chronic desmoplastic pancreatitis or involved by a malignant
tumor are not likely to be dilated by secretin. Thus,
the use of secretin may increase the accuracy of
MRCP in detecting pathological stenosis of the
MPD. Secretin administration is very effective in
depicting side branches and the undilated MPD.
One clinical unit/kg of secretin is good for dilating
the pancreatic duct, and in a delayed phase, it can,
to some extent, provide information about excretory pancreatic function 15,45,46(Fig 3A and B). The
MPD is not always entirely visualized when secretin is not used. With secretin, MRCP can depict the
biliopancreatic bifurcation more clearly. Anatomic
variants like pancreas divisum are also detected
with a high confidence level. Even with secretin
administration, however, delineation of side
branches is rare in normal individuals. Because
secretin inhibits gastrin effect, the administration of
secretin can increase the duodenal fluid without
increasing gastric juice secretion. This effect is
beneficial in evaluating the pancreatic ducts and
duodenal and periampullary lesions without the
problematic overlap of the gastric juices.
Taking advantage of the endoscopy-independent
approach of MRCR postsurgical evaluation of the
pancreatic duct has previously been reported. 47
Secretin administration also is useful in evaluating
postsurgical assessment of the exocrine pancreatic
function or patency of the pancreaticojejunostomy48 (Fig 4A and B). According to Sho et al, 48 in
their series of patients after pancreaticoduodenectomy, a positive correlation was shown between
oral glucose tolerance test and MRCP grade determined by the amount of exocrine pancreatic secretion after secretin administration.

Background Suppression
In MRCP, any water signals originating from
other than the pancreaticobiliary system become
background noise. Intravenous administration of
gadolinium chelate is useful in suppressing the

328

YASUO TAKEHARA

Likewise, ingestion of ferric ammonium citrate

as a negative contrast agent also is useful for
suppression of background signals within the gastrointestinal tracts. 5

Anticholinergic Drugs
Suppression of bowel motion is essential in
using MRCP based on multishot FSE or 3DFT FSE
with lengthy imaging times. When peristalsis of the
fluid-filled bowel occurs during imaging, it creates
considerable artifacts in the phase encoding direction. We use 7.5 mg of timepidium bromide, when
indicated. With the use of snap shot imaging,
however, combined use of anticholinergic drugs
may not be mandatory because they interfere with
observation of physiological bowel motion, peristalsis of the Oddi's sphincter, and so on. Also, the use
of anticholinergic drugs may potentially affect
pancreatic exocrine secretion tested by secretin,
even though the output of which is basically
hormonally regulated rather than vagal.
Our routine imaging parameters are shown in
Table 1.

CLINICAL CONSIDERATIONS

Normal Pancreatic Duct

Fig 3. A 79-year-old man with a side branch dilatation. (A)

Before secretin administration, the MPD is not well delineated
on single thick slab MRCR (B) Fifteen minutes after secretin
administration, a single thick slab MRCP shows that the MPD
is fully delineated. MPD was normal, except for its coiling at
the head portion (large arrow) and the side branch dilatation
in the tail portion (small arrow).

background signals in MRCP images. Single thick

slice method uses heavily T2-weighted images to
avoid signals from medium T2 substances; however, there is still some overlap of signals created
by slow flowing blood, free water in the interstitium, and urine in the urinary tract, and so on.
Therefore, the T2 shortening effect obtained by
paramagnetic contrast media that is distributed in
the vessels, interstitium, and urinary tract effectively suppresses the background signals. Thus, a
single dose of gadolinium chelate enables improved CNR on the acquired MRCP 49 (Fig 5A
and B).

The pancreas develops from endodermal outgrowths of the foregut. It consists of 2 components
that are called the ventral pancreas and the dorsal
pancreas. From the sixth week of development,
ducts form in each pancreatic component and
communicate with the foregut lumen. The ventral
duct drains in common with the bile duct through
the developing major papilla; the dorsal duct drains
slightly cephalad through the minor papilla. As the
parenchymal components of the pancreas fuse,
fusion of ductal systems occurs in most people.
Fusion typically occurs in the pancreatic neck with
the dorsal duct draining the tail and body of the
pancreas, whereas the accessory duct of Santorini
drains distally from the site of fusion. The Santorini
duct normally communicates with the duodenum
through the minor papilla, but, in some cases, it
may regress and, therefore, lose its direct communication. The duct of the ventral pancreas (Wirsung's
duct) forms the continuation of the MPD, with an
extention from the site of fusion to the papilla of
Vater. These ducts form the main streams of
pancreatic drainage. 51-54

MR PANCREATOGRAPHY

329

Fig 4. A 45-year-old man after removal of papilla of Vater cancer. The patient underwent pancreaticoduodenectomy and pancreaticojejunostomy. Secretin MRCP is useful in evaluating the patency of the anastomosed channel as well as exocrine function reserve of the pancreas
to some extent. (A) Before secretin administration, the depiction of the MPD is unsatisfactory. (B) After secretin, the tail segment of the MPD
is successfully opacified (arrows). Also note there is an increase of pancreatic juice that empties into the jejunal loop (J).

Anomalies

Pancreas divisum is the most common congenital anomaly of the human pancreas. It occurs when
the ventral and dorsal parts of the pancreas fail to
fuse resulting in pancreatic drainage via the Santorini duct through the accessory papilla. 55 Therefore, the combination of pancreas divisum and a
small accessory orifice could result in dorsal duct
obstruction. The challenge is to identify dorsal duct
pathologic conditions in this group of patients. 55-57
Previous documentation of this anomaly was largely
dependent on ERCE Because cannulation of the
dorsal duct by ERCP is more difficult than cannulation of the Wirsung's duct, it is likely that many
patients with this anomaly might have been overlooked. Also, it should be stressed that with ERCP,
the appearance of pancreas divisum, ie, a smallcaliber ventral duct with an arborizing pattern, may
be confused with an obstructed MPD secondary to
pancreatic cancer.
MRCP can detect this anomaly together with
accompanied ductal lesions 56 (Fig 6A and B). The
key sign of pancreatic fusion anomaly is that a
dominant pancreatic duct can be tracked continuously from the tail to the head portion, where it
crosses anterior to the common bile duct and
empties into the duodenum at the accessory papilla.
This finding is better appreciated using multiplanar
reformation rather than projection reconstruction.
This anomalous course of the pancreatic duct is
more positively documented when secretin is administered. 45,58In our experiences with secretin-MRCP,
delineation of the pancreatic duct is even more

Fig 5. A 67-year-old man suspected of having chronic

pancreatitis. (A) Before administration of single dose of gadolinium chelate (Gd-DTPA-BMA), background signals are somewhat problematic in evaluating details of the pancreaticobiliary system. MRCP acquired with a single thick slab method.
(B) After intravenous gadolinium injection, background signals are successfully suppressed on the MRCP taken with the
same technique as in Fig 5A above.

330

YASUO TAKEHARA
Table 1, Routine MRCP Method in our Institution

NEX

Matrix

FOV (cm)

Options

Coronal T2~:
Coronal T2

2D SSFSE
2D SSFSE

Sequence

=/600-1,000/90
~0/80-100/90

TR/TE/FA

0.5
0.5

256 x 160-256
256 160

20 20
24 18

Axial T2

2D SSFSE

oo/80-100/90

0.5

256 160

24 x 18

Single thick slice (30-50 mm)t, non-BH

Multiple thin slices (3-5 ram), interleave, resp
gated, CHESS ( - )
Multiple thin slices (3-5 mm), interleave, resp
gated, CHESS ( - )

NOTE. The routine echo space is 4.5 milliseconds at receiver band-width of 62 KHz. TR and TE are in milliseconds.
Abbreviations: TR, repetition time; FA, flip angle (degrees); CHESS, chemical shift selection; NEX, number of excitations; Non-BH,
non-breath-held; resp gated, respiratory gated.
tMultiangle oblique MRCP is done if available (slab thickness, 20 mm). (A single dose of Gd-DTPA may be administered before
imaging for background signal suppression.)
CFor pharmacodynamic approach, the sequence is repeated with smaller matrix size of 256 x 128-160 and shorter TE of 200 ms
after 1 CU/kg secretin.

straightforward in divisum patients than in normal

individuals. In patients with pancreas divisum,
increasing pancreatic juice flow stimulated by
secretin may exceed the draining capacity of the
minor papilla into the duodenum.
Bret et aP 6 reported that MRCP diagnosed
pancreas divisum as accurately as ERCP did. Their
study was conducted using ERCP as the gold
standard. Theoretically, MRCP can identify more
pancreas divisum than ERCR Unlike ERCP, the
detection of divisum with MRCP is not affected by
technical failure of endoscopic insertion or cannulation. As Bret et al implied in their study, the
incidence of pancreas divisum using MRCP with
torso array coil and conventional FSE sequence
was 12% in their population. This rate was much
higher than the previously reported rate based on
ERCP studies; however, it was closer to the rate
reported in autopsy series. 54,56
Basically, patients with pancreatitis and pancreas
divisum should be treated with conservative measures, including pancreatic enzyme therapy. Many
of these patients have idiopathic pancreatitis unrelated to the pancreas divisum and will respond well
to the conservative therapy. However, surgical
intervention is indicated when the above attempts
fail. If marked dilation of the dorsal duct can be
shown, ductal decompression surgery is recommended.57,59

Adenocarcinoma of the Pancreas

Pancreatic adenocarcinoma is an important clinical problem, accounting for 5% of all cancerrelated deaths each year, and is the fourth leading
cause of death by cancer in the United States. With
the exception of jaundice and long history of
diabetis mellitus, 6 the initial symptoms associated

with pancreatic cancer are often insidious in nature

and usually precede diagnosis by longer than 2
months. In spite of the advancement of serologic
tests for tumor-associated antigens such as carcinoembryonic antigen and carbohydrate antigen 19-9
and detection of K-ras mutation in endoscopically
sampled pancreatic juice, 61 as well as development
of noninvasive imaging techniques such as CT and
US, the early diagnosis of a potentially resectable
pancreatic carcinoma remains extremely difficult.
In Japan, the detection of early pancreatic cancer
(stage I) is disappointingly low (6.9%). 62 Recent
MR approaches using fat-saturated, Tl-weighted
images with gadolinium chelate and dynamic contrast enhanced spiral CT may improve the accuracy
of pancreatic cancer diagnosis: however, even
small tumors are often accompanied by wide
spread to the retroperitoneum. In situations where
additional diagnostic evaluation is required, ERCP
may clarify the cause of ambiguous CT or US
findings. The characteristic findings are stenosis or
obstruction of either the pancreatic or the common
bile duct. Both duct systems are abnormal in over
half of the cases. Distinguishing between carcinoma and chronic pancreatitis by ERCP is often
quite difficult because some pancreatic carcinomas
coexist with pancreatitis. Both pancreatitis and
pancreatic carcinoma can cause enlargement and
produce similar-appearing changes in the pancreas
and peripancreatic tissues.
Symptomatic pancreatic cancer usually involves
the MPD, so MRCP can depict the narrowed or
amputated MPD as well as dilated upstream ducts
quite well (Fig 7A and B). Ampullary cancer is
shown by diffuse and smooth dilatation of the
biliary duct and the pancreatic duct. With adequate
fluid filling, the tumor is often outlined within the

MR PANCREATOGRAPHY

331

scopic retrograde pancreatography (ERP), particularly balloon occluded ERR seems to be more
sensitive than MRCP to detect small eccentric
cancers. Recently, the ability of endoscopic ultrasonography (EUS) to detect small pancreatic cancer
also has been reported.
To date, the pattern of affected pancreatic ducts,
such as irregularity or duct penetrations, contributes little to the differential diagnosis of pancreatic
diseases. However, large masses of the pancreatic
head region not causing obstruction of the ducts
usually have a histologic condition other than
ductal cancer, including metastases or lymphoma
in the peripancreatic lymph nodes or focal pancreatitis. If spatial resolution of MRCP is further
improved, it may provide useful information on the
status of the pancreatic ductal system and thus aid
in the differential diagnosis of pancreatic disease.
Parenchymal Contrast and Contrast Enhancement

Fig 6. A 90-year-old man with chronic pancreatitis associated with pancreas divisum. (A) A single thick slab MRCP
shows diffusely dilated dorsal pancreatic duct continuing to
the duct of Santorini (small arrow). There are also some filling
defects within the MPD (large arrows), but the diagnosis of
pancreatic calculi is inconclusive. (B) Multiple thin slice with
medium TE without fat saturation can depict oval pancreatic
calculi within the MPD (arrows). With this method, the boundary of the pancreas as well as the ductal wall is demarcated.

duodenum around the papilla of Vater. Nevertheless, ERCP is superior to MRCP in the direct
observation of the duodenal mucosa such as color
and textures acquired as magnified images of the
mucosal surface as well as the evaluation of the
ampullary region, combining the endoscopic view
with contrast radiography.
MRCP detection of early pancreatic cancer without MPD involvement is difficult. Because of its
ability to depict side branches and parenchymal
enhancement (or acinar contrast filling), endo-

In addition to ductal imaging, the parenchymal

contrast derived from Tl-weighted images is particularly important in evaluating pancreatic cancer.
It is virtually impossible to make a diagnosis of
pancreatic cancer using hydrographic images alone.
Pancreatic carcinoma is depicted as a low-intensity
lesion surrounded by high signal of intact pancreatic parenchyma. In patients with chronic pancreatitis, however, baseline pancreatic signal intensity on
Tl-weighted images may be reduced so that tumors
are often inconspicuous. Likewise, with advancing
age, the normal pancreas signal intensity may
diminish to equal that of the liver. It is, therefore,
important to use dynamic contrast enhanced MR
imaging with bolus injection of gadolinium chelate
in conjunction with Tl-weighted gradient-echo
sequence to produce parenchymal contrast63 (Fig
7C and D).
Gadolinium chelate injection followed immediately by dynamic imaging, such as with breathholding Tl-weighted gradient echoes, is the most
reliable method of detecting small pancreatic carcinomas with MRI. The pancreas is enhanced before
the liver after bolus intravenous administration of
gadolinium chelate. Maximal enhancement of the
pancreas occurs soon after peak aortic enhancement and remains high during the late portal phase
of liver enhancement. Ductal cancers appear distinctly hypointense compared with the pancreas
during the first 1 to 2 minutes after injection, when
the pancreas is maximally enhanced. Dynamic

332 "

YASUO TAKEHARA

Fig 7. A 6g-year-old woman with a cancer in the pancreatic body. (A) ERCP could not fill contrast media upstream to the
obstruction (black arrow) in the body portion of the MPD, (B) MRCP showed entire pancreatic duct together with the severe stenosis
(white arrow) in the body portion. (C) Contrast enhanced Tl-weighted image acquired with 3DFT fast spoiled GRASS sequence after
bolus injection of Gd-DTPA-BMA (arterial phase), There is a poorly enhanced mass (arrow) measuring lcm in the pancreatic body.
The upstream pancreatic duct is diffusely dilated. (D) On the portal phase contrast enhanced T1-weighted image acquired with the
same technique, there is a gradual contrast filling in the mass, but is still less opacified as compared with the normal pancreas.

contrast-enhanced MR is competitive with dynamic CT studies using helical CT, 64'65

Mucinous Cystic Neoplasia

Intraductal mucin-producing tumors (MPT) originate from the epithelium of the MPD or side
branches. The tumor is characterized by production
of a large amount of mucin. Excess mucin production can cause biliary obstruction by a mechanical
compression or by biliary congestion caused by a

large amount of mucin that empties into the biliary

duct through a fistula. MPT comprises roughly 2%
of nonendocrine pancreatic carcinomas. 66-7 Histologically, there are large cystic spaces filled with
mucin and surrounded by connective tissue septa. 71
The MPT limited to the ducts consists of flat or
polypoid tumors confined to the MPD and its side
branches. The pancreatic ducts are dilated and filled
with the mucin produced by the tumor. Rather than
being a separate entity, these may represent a less

MR PANCREATOGRAPHY

333

Fig 8. A 70-year-old man with MPT in the pancreatic tail.

(A) Hindered by mucinous fluid, ERP could not fill the contrast
media into the lesion. Therefore, the cystic neoplasm could
not be seen in this particular image. (B) Contrarily, MRCP
acquired with single thick slab showed a cystic lesion (arrow)
with clear continuity with the dilated MPD in the tail portion.
The image was taken while the patient exhaled. (C) MRCP
acquired with the same method during inhalation successfully
depicted the cystic lesion (arrow) in the tail portion avoiding
overlap with the gastric juice.

aggressive mucinous cystadenocarcinoma diagnosed at a relatively earlier stage. Unlike other

ductal cancers, the outcome of surgical resection
for this neoplasm is favorable, hence early diagnosis of this tumor is critical. 69
In a study using conventional imaging, Obara et
al72 reported that all cases involving the MPD were
malignant (n = 3), and that nearly 90% (8 of 9) of
the branch duct type were benign. Intraductal
papillary lesions larger than 6 m m were either
adenomas or adenocarcinomas, and they found that
MPT with a papillary lesion smaller than 6 mm
could be followed up carefully. On ERCE diffuse
or segmental dilatation of the MPD or cystic
dilatation of the side branches, sometimes grapelike, are common. Fungating lesions are sometimes
detected as filling defects in ERCE 73 Because
ERCP can detect the communicating channels
between the cystic lesions and the MPD, ERCP is
said to be the most reliable method in diagnosing
this neoplasm; however, the mucin often impedes
contrast filling v (Fig 8A). With use of MRCR on
the other hand, detection of communicating chart-

nels as well as depiction of the entire cystic lesion

is not impeded by mucin at all; thus, a more
accurate diagnosis of this predominantly cystic
tumor is feasible 74,75 (Fig 8B and C).
MRCP and EUS are currently the best modalities
for determining the size of papillary lesions within
the cystic lesions as well as the type of ductal
dilatation, either a branch duct or main duct. 74'75
Because papillary outgrowth on the surface of the
cystic wall manifests as a filling defect, the solid
lesion is measurable. Because of its noninvasiveness, the use of MRCP for MPT follow-up might be
feasible.
Chronic Pancreatitis
In chronic pancreatitis, ERCP can depict ductal
abnormalities including luminal narrowing, irregularities in the ductal system such as stenosis,
dilatation, sacculation, and ectasia, and blockage of
the pancreatic duct by calcium deposits. Differentiation from carcinoma may be difficult because of
overlapping features such as stenosis and irregularity of the ducts.

334

MRCP is useful in depicting dilated or narrowed

segments of the MPD, which are characteristic of
chronic pancreatitis. 19Pancreatic calculi and plaque
are detectable as round filling defects within the
MPD 19,20,27 (Fig 6A and B). Similar!y, bubbles
within the dilated pancreatic duct may result in
false-positive findings, 76 therefore, special care
should be taken in evaluating the intraductal lesions
of patients with MPT or individuals who underwent
sphincterectomy. Deviation and/or dilatation of the
biliary duct, when involved, is also seen. 18-2,z7
Typically, the side branches of the MPD are not
depicted unless dilated.
One of the important radiographic hallmarks of
chronic pancreatitis is the presence of scattered
calcification throughout the pancreas. Diffuse pancreatic calcification indicates that significant damage has occurred. Calcification is especially prominent in alcohol-induced pancreatitis, but it also may
be seen in other abnormalities such as high protein
malnutrition, hereditary pancreatitis, posttranmatic
pancreatitis, hyperparathyroidism, islet cell tumors,
and idiopathic chronic pancreatitis. According to
Ammann et al, 77 pancreatic calcification may decrease or even disappear spontaneously after ductal
decompression. This dynamic behavior of pancreatic calcification seen in chronic pancreatitis may
be used as an indicator of the advancement of the
disease. One disadvantage of MR may be its poor
sensitivity to detect such calcifications.
The role of MRCP in advanced cases of chronic
pancreatitis may be: first, the evaluation of ductal
pathologic conditions, which may benefit planning
of surgical or endoscopic interventions and, second, follow-up after said interventions. When the
pain induced by chronic pancreatitis is severe
enough to require frequent use of analgesics and/or
narcotics, an appropriate surgical intervention
should be considered to alleviate pain. MRCP
allows the surgeon to plan the surgical approach. If
there is a stricture of the pancreatic duct, then a
local resection may relieve the pain. Unfortunately,
isolated localized stenoses are uncommon. In most
patients with alcohol-induced pancreatitis, the pancreatic ducts are almost always diffusely affected
and are difficult to be corrected surgically. When
there is primary ductal obstruction, as caused by
pancreatic calculi, surgical decompression may
provide effective pain palliation. Before surgery,
precise determination of the obstructive point is
essential because patients are prone to endocrine

YASUO TAKEHARA

and exocrine insufficiency and must be treated with

pancreatic enzyme replacement. As for nonsurgical
interventions, endoscopic placement of stents into
the pancreatic duct has been attempted. In such
cases, MRCP plays an important role because only
MRCP can precisely locate the stenotic segment.
ERCP usually fails to fill upstream of the severe
stenosis with contrast media, whereas MRCP can
disclose the exact length of the stenotic segment in
the MPD. Likewise, the outcome of endoscopic or
extracorporeal shockwave lithotripsy78can be evaluated by MRCE
In addition to excluding pseudocysts and pancreatic cancer, CT and US may show the pancreatic
stones or ductal dilatation associated with chronic
pancreatitis. The next step should be an evaluation
of ductal lesions. ERCP provides a detailed view of
the pancreatic ducts. In patients with alcoholinduced pancreatitis, ERCP may show a pseudocyst missed by sonography or CT scan. Because of
its noninvasive nature, MRCP is even more useful
for detecting pancreatic pseudocysts, imaging ducts,
and investigating benign and chronic abnormalities. MRCP also is a reasonable choice for follow-up studies.
The diagnosis of early chronic pancreatitis is
currently very difficult in many cases. ERP is
indicated when other tools fail to prove pancreatic
diseases; however, even ERP often fails to show
any evidence of pancreatic diseases. With future
improvements in the spatial resolution of MRCP as
well as the combined use of secretin, subtle changes
of side branches might be detected; thereby allowing diagnosis at an early stage of segments affected
by chronic pancreatitis.
Parenchymal contrast acquired with fat saturated
Tl-weighted images also is helpful in determining
the affected area. Furthermore, this technique,
combined with the use of gadolinium chelate
contrast media, may increase the reliability of
diagnosing pancreatic inflammatory processes. The
areas affected by inflammatory processes show
decreased enhancement in the arterial phase, and
increased enhancement in the equilibrium phase as
opposed to the normal pancreatic segments. 63,64
Acute Pancreatitis

CT and US are currently the initial imaging

modalities used to visualize the pancreas and
peripancreatic spaces, but are often inconclusive
for staging acute pancreatitis and detecting compli-

MR PANCREATOGRAPHY

cations. When pancreatic duct leakage is suspected

in the presence of pancreatic ascites, ERCP is
currently used to detect and localize its site and for
the choice of appropriate therapy. ERCP may show
contrast leakage from the pancreatic duct into the
fluid collection. However, injection of contrast
media into a cyst may cause infection, resulting in
pancreatic or peripancreatic abscesses. 79 MRCP is
safe to use for providing information about: (1) the
presence or absence of ductal distention, (2) disruption, (3) leakage of the pancreatic duct, and (4)
intraductal lesions that might increase intraductal
pressure resulting in a predisposition to pseudocyst
formation. In addition to the detection and localization of pseudocysts, MRCP also is straightforward
in determining their size, shape, and number. This
information is important because it determines the
severity of the acute pancreatitis and influences the
choice of treatment.
Emergency ERCP is indicated in patients with
suspected biliary pancreatitis to remove possible
impacted gallstones at the ampulla. 8 MRCP can
indirectly diagnose gallstone pancreatitis by depicting a diffusely dilated MPD and biliary duct. The
impacted gallstone itself, however, might elude
detection with MRCP because biliary or pancreatic
fluid might not surround a tightly impacted stone.
CT may help detect the impacted stone at the
ampulla by its inherent sensitivity to calcification if
the stone is calcified.
In addition, determination of the area of involvement by peripancreatic inflammation and phlegmon should be made based on conventional T2weighted images rather than MRCE

Kinematic Approach Using Single Thick Slice With

Long TE
After the administration of 1 CU/kg of secretin,
coronal imaging is repeated every 4 to 6 seconds
for about 8 minutes. 29 Our standard parameters use
a single thick slab image taken in a coronal plane
with TE of 200 to 600 milliseconds, a thickness of
20 to 50 mm, a field of view (FOV) of 20 X 20 cm,
and a matrix of 256 X 128 to 160. Generally, a
smaller phase matrix is used in kinematic approach
compared with ordinary MRCP to maintain the
brief imaging time of the individual images. This is
analogous to a "high shutter speed camera" which
can freeze the objects' movement. The individual
images are observed on the display simultaneously,
and the series of images are also appreciated from a

335

kinematic viewpoint using cine loop. This method

is what we term pharmacodynamic MRCP, and, if
done without secretin administration, it is called
kinematic MRCP. Strictly speaking, dynamic behavior of the organs during kinematic MRCP is not
real-time information. However, a temporal resolution of 4 seconds can follow the peristalsis or slow
respiratory movement of the abdominal organs.

Kinematic Evaluation of the Oddi's Sphincter and

Biliary Dyskinesia
A higher than expected number of patients suffer
from unexplained abdominal pain that may resemble a symptom of pancreatitis. Likewise, some
patients who underwent cholecystectomy also complain of recurrent pain. After a series of noninvasive examinations, ERCP has traditionally been
used to identify the cause of such pain. If no lesion
is found within the biliary and/or pancreatic ducts
in a patient with such repeated attacks of acute
pancreatitis, manometric studies of the Oddi's
sphincter may be indicated. Such studies, however,
do increase the risk of post-ERCP/manometry
acute pancreatitis. Also, the risks of allergic reactions to contrast media and premedications as well
as the hazards of ionizing radiation may not be
justified to diagnose this likely benign condition.
Kinematic studies using fast MR technologies
will probably become prevalent for the depiction of
dynamic processes that will show more physiological and pathological information about this field
(Fig 9A and B). Dynamic, or kinematic, MR studies
of the pancreaticobiliary system have important
potential for identifying and characterizing abnormalities originating in the corresponding areas.
MRCP with ultrafast imaging techniques such as
the single thick slice method with long TE SSFSE
can freeze the motion and allows kinematic observation of the dynamic processes. With ordinary
MRCP that requires 2 seconds or more for imaging,
dynamic processes of the biliopancreatic structures
may be missed, or merely cause artifacts that may
be misleading.
Another potential of ultrafast MRCP is its increased use for interactive procedures such as the
guidance of biopsies and intravascular procedures,
as well as for the monitoring of ablations. Realtime visualization of the dynamics and interactive
scan control during continuous scanning as well as
the capability to track minimally invasive devices

336

YASUO TAKEHARA

are essential for the fulfillment of MR as an

interventional tool.

Cost Effectiveness and Indications of MRCP

In the current cost-conscious healthcare environment, MRCP undoubtedly decreases medical expenditures. Highly advanced cases of pancreatic cancer
may not require ERCE MRCP is cheaper and more
efficient to overview the pancreaticobiliary system
noninvasively. Furthermore, because of its noninvasiveness and lower cost, MRCP may replace ERCP
for screening purposes or just for follow-up investigation. ERCP will gradually be limited to interventional purposes only.

MRCP Accessibility: A Common Concern

An important question, though, is how accessible MR is in the current clinical environment. In
many Japanese medical institutions, MR examinations are usually fully booked. This is especially
relevant to high performance imagers. Conversely,
ERCP may be scheduled earlier and with greater
flexibility. The order of examinations in the decision tree may thus be reversed in spite of MRCP's
economic effectiveness and patient comfort. Thus,
the use of MR for screening patients may be
relatively limited. Ideally, a flexible booking system with a standardized MRCP protocol would
allow the early incorporation of MRCP into the
diagnostic decision tree. Of course, the screening
protocol should be performed by sequences with
good quality control, otherwise, false lesions created by poor MRCP examinations may increase
unnecessary ERCP examinations or surgical interventions.

Fig 9. An 80-year-old woman who underwent cholecystectomy 30 years ago complained of recurrent pain in the
abdomen. Because US showed diffusely dilated common
biliary duct, a cancer of the papilla of Vater was suspected.
Pharmacodynamic MRCP using single thick slab method and
secretin administration was performed. Observe the function
of the Oddi's sphincter. Observation of dynamic images focusing on the Oddi's sphincter confirmed that the function was
maintained. No tumor was identified at the corresponding
area at the same time, therefore, the dilatation of the choledochus was considered to be a benign, nonspecific dilation after
cholecystectomy. (A) MRCP taken when the Oddi's sphincter
relaxed. The channels of biliary duct and pancreatic duct are
open {see inset). (B) MRCP obtained when the channels were
closed (see inset).

MR PANCREATOGRAPHY

337

REFERENCES
1. Keogan MT, McDermott VG, Paulson EK, et al: Pancreatic malignancy: Effect of dual-phase helical CT in tumor
detection and vascular opacification. Radiology 205:513-518,
1997
2. Campbell JP, Wilson SR: Pancreatic neoplasms: How
useful is evaluation with US? Radiology 167:341-344, 1988
3. Huntington DK, Hill MC, Steinberg W: Biliary tract
dilatation in chronic pancreatitis: CT and sonographic findings.
Radiology 172:47-50, 1989
4. White Jr RI, Halden Jr WJ: Liquid gold: Low-osmolality
contrast media. Radiology 159:559-560, 1986
5. Wolf GL: Safer, more expensive iodinated contrast agents:
How do we decide? Radiology 159:557-558, 1986
6. Barrett BJ, Parfrey PS, McDonald JR, et al: Nonionic
low-osmolality versus ionic high-osmolality contrast material
for intravenous use in patients perceived to be at high risk:
Randomized trial [see comments]. Radiology 183:105-110,
1992
7. Yamaguchi K, Katayama H, Takashima T, et al: Prediction
of severe adverse reactions to ionic and nonionic contrast media
in Japan: Evaluation of pretesting. A report from the Japanese
Committee on the Safety of Contrast Media. Radiology 178:363367, 1991
8. Kang JK, Chung JB, Moon YM, et al: The normal
endoscopic pancreatogram in Koreans. Korean J Intern Med
4:74-79, 1989
9. Cohen SA, Siegel JH, Kasmin FE: Complications of
diagnostic and therapeutic ERCR Abdom Imaging 21:385-394,
1996
10. Sherman S, Lehman GA: ERCP- and endoscopic sphincterotomy-induced pancreatitis. Pancreas 6:350-367, 1991
11. Kaneko E, Harada H, Kasugai T, et al: Report of the
second nationwide survey of the complications related to
gastroenterological endoscopy. [Japanese] Gastroenterol Endosc 37:642-652, 1995
12. Cohen RV, Aldred MA, Paes WS, et al: How safe is
ERCP to the endoscopist? Surg Endosc 11:615-617, 1997
13. Outwater EK, Gordon SJ: Imaging the pancreatic and
biliary ducts with MR. Radiology 192:19-21, 1994 (editorial;
comment)
14. Jara H, Barish MA, Yucel EK, et al: MR hydrography:
Theory and practice of static fluid imaging. AJR Am J Roentgenol 170:873-882, 1998
15. Takehara Y, Ichijo K, Masni T, et al: Secretin MRCP:
Improved MPD delineation and the potential for evaluating
pancreatic exocrine function reserve. Proceedings of International Society for Magnetic Resonance in Medicine 1996;802.
16. Morimoto K, Shimoi M, Shirakawa T, et al: Biliary
obstruction: Evaluation with three-dimensional MR cholangiography. Radiology 183:578-580, 1992
17. Wallner BK, Schumacher KA, Weidenmaier W, et al:
Dilated biliary tract: Evaluation with MR cholangiography with
a T2-weighted contrast-enhanced fast sequence. Radiology
181:805-808, 1991
18. Laubenberger J, Buchert M, Schneider B, et al: Breathhold projection magnetic resonance-cholangio-pancreatography
(MRCP): A new method for the examination of the bile and
pancreatic ducts. Magn Reson Med 33:18-23, 1995
19. Takehara Y, Ichijo K, Tooyama N, et al: Breath-hold MR
cholangiopancreatography with a long-echo-train fast spin-echo

sequence and a surface coil in chronic pancreatitis [see comments]. Radiology 192:73-78, 1994
20. Barish MA, Yucel EK, Soto JA, et al: MR cholangiopancreatography: Efficacy of three-dimensional turbo spin-echo
technique [see comments]. AIR Am J Roentgenol 165:295-300,
1995
21. Macaulay SE, Schulte SJ, Sekijima JH, et al: Evaluation
of a non-breath-hold MR cholangiography technique. Radiology
196:227-232, 1995
22. Miyazaki T, Yamashita Y, Tang Y, et al: Single-shot MR
cholangiopancreatography of neonates, infants, and young children. AJR Am J Roentgenol 170:33-37, 1998
23. Pavone P, Laghi A, Catalano C, et al: MR cholangiopancreatography (MRCP) at 0.5 T: Technique optimisation and
preliminary results. Enr J Radiol 6:147-152, 1996
24. Reinhold C, Guiband L, Genin G, et al: MR cholangiopancreatography: Comparison between two-dimensional fast spinecho and three-dimensional gradient-echo pulse sequences. J
Magn Reson Imaging 5:379-384, 1995
25. Guibaud L, Bret PM, Reinhold C, et al: Diagnosis of
choledocholithiasis: Value of MR cholangiography. AJR Am J
Roentgenol 163:847-850, 1994
26. Pavone P, Laghi A, Panebianco V, et al: MR cholangiography: Techniques and clinical applications. Eur J Radiol
8:901-910, 1998
27. Soto JA, Barish MA, Yucel EK, et al: Pancreatic duct:
MR cholangiopancreatography with a three-dimensional fast
spin-echo technique. Radiology 196:459-464, 1995
28. Roemer PB, Edelstein WA, Hayes CE, et al: The NMR
phased array. Magn Resort Med 16:192-225, 1990
29. Takehara Y: Fast MR imaging for evaluating the pancreaticobiliary system. Eur J Radio129:211-232, 1999
30. Hennig J, Nanerth A, Friedburg H: RARE imaging: A fast
imaging method for clinical MR. Magn Reson Med 3:823-833,
1986
31. Hennig J, Nauerth A, Friedburg H, et al: New rapid
imaging procedure for nuclear spin tomography. Radiologe
24:579-580, 1984
32. Mulkern RV, Wong ST, Winalski C, et al: Contrast
manipulation and artifact assessment of 2D and 3D RARE
sequences. Magn Reson Imaging 8:557-566, 1990
33. Oshio K, Jolesz FA, Melki PS, et al: T2-weighted
thin-section imaging with the multislab three-dimensional RARE
technique. J Magn Reson Imaging 1:695-700, 1991
34. Barish M, Soto J, Ferrucci J: Magnetic resonance pancreatography. Endoscopy 29:487-495, 1997
35. Barish MA, Soto JA, Yucel EK: Magnetic resonance
cholangiopancreatography of the biliary ducts: Techniques,
clinical applications, and limitations. Top Magn Reson Imaging
8:302-311, 1996
36. Miyazaki T, Yamashita Y, Tsuchigame T, et al: MR
cholangiopancreatography using HASTE (half-Fourier acquisition single-shot turbo spin-echo) sequences. AIR Am J Roentgenol 166:1297-1303, 1996
37. Takehara Y: Can MRCP replace ERCP? J Magn Reson
Imaging 8:517-534, 1998
38. 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

338

multislice technique. AJR Am J Roentgenol 168:1449-1454,

1997
39. McDermott VG, Nelson RC: MR cholangiopancreatography: Efficacy of three-dimensional turbo spin-echo technique
[comment]. AJRAm J Roentgenol 165:301-302, 1995
40. Elster AD: Creating an MR image, in Questions and
Answers in Magnetic Resonance Imaging. St Louis, MO,
Mosby-Year Book, 1994, pp 109-110
41. Drebin RA, Carpenter L, Hanrahan P: Volume rendering.
Computer Graphics 1988;22:65-74
42. Constable RT, Gore JC: The loss of small objects in
variable TE imaging: Implications for FSE, RARE, and EPI.
Magn Reson Med 28:9-24, 1992
43. Sananes J, Bonnet M, Lecesne R, et al: Magnetic
resonance cholangiography using HASTE sequence. Proceedings of the Society of Magnetic Resonance 1995;1453
44. Mizutani Y, Ichikawa T, Hanaoka H, Haradome H,
Katase S, Hachiya J: MR-cholangiopancreatography using
HASTE sequence: Comparison with fast-SE and CE-FAST.
[Japanese] Jpn J Magn Resort Med 15:277, 1995 (abstr)
45. Matos C, Metens T, Deviere J, et al: Pancreatic duct:
Morphologic and functional evaluation with dynamic MR
pancreatography after secretin stimulation. Radiology 203:435441, 1997
46. Takehara Y, Ichijo K, Tooyama N, et al: Enhanced
delineation of the pancreatic duct in MR cholangiopancreatography (MRCP) with a combined use of secretin. [Japanese].
Nippon Igaku Hoshasen Gakkai Zasshi 55:255-256, 1995
47. Yamaguchi K, Chijiwa K, Shimizu S, et al: Comparison
of endoscopic retrograde and magnetic resonance cholangiopancreatography in the surgical diagnosis of pancreatic diseases.
Am J Surg 175:203-208, 1998
48. Sho M, Nakajima Y, Kanehiro H, et al: A new evaluation
of pancreatic function after pancreatoduodenectomy using secretin magnetic resonance cholangiopancreatography. Am J Surg
176:279-282, 1998
49. Asakura K, Ooshima T, Ikeda K, et al: The value of
Gd-DTPA MRCP in medium field MRI. [Japanese] Nippon
Igaku Hoshasen Gakkai Zasshi 58:249, 1998 (abstr)
50. Hirohashi S, Hirohashi R, Uchida H, et al: MR cholangiopancreatography and MR urography: Improved enhancement
with a negative oral contrast agent. Radiology 203:281-285,
1997
51. Meschan I: The pancreas, in An Atlas of Anatomy Basic
to Radiology. Philadelphia, PA, Saunders, 1975, pp 823-833
52. Kleitsch WP: Anatomy of the pancreas. A study with
special reference to the duct system. Arch Surg 71:795-802,
1955
53. Dawson W, Langman J: An anatomical-radiological
study on the pancreatic duct pattern in man. Anat Rec 139:5968, 1961
54. Berman LG, Prior JT, Abramow SM, et al: A study of the
pancreatic duct system in man by the use of vinyl acetate casts of
postmortem preparations. Surg Gynecol Obstet 110:391-403,
1955
55. Gregg J, Solomon J, Clark G: Pancreas divisum mad its
association with choledochal sphincter stenosis. Diagnosis by
endoscopic retrograde cholangiopancreatography and endoscopic biliary manometry. Am J Surg 147:367-371, 1984
56. Bret PM, Reinhold C, Taourel P, et al: Pancreas divisum:

YASUO TAKEHARA

Evaluation with MR cholangiopancreatography. Radiology 199:

99-103, 1996
57. O'Rourke RW, Harrison MR: Pancreas divisum and
stenosis of the major and minor papillae in an 8-year-old girl:
Treatment by dual sphincteroplasty. J Pediatr Surg 33:78%791,
1998
58. Takehara Y: MR pancreatography: Technique and applications. Top Magn Reson Imaging 8:290-301, 1996
59. Rusnak CH, Hosie RT, Kuecbler PM, et al: Pancreatitis
associated with pancreas divisum: Results of surgical intervention. Am J Surg 155:641-643, 1988
60. Everhart J, Wright D: Diabetes mellitus as a risk factor
for pancreatic cancer. A meta-analysis. JAMA 273:1605-1609,
1995
61. Okada S, Yoshimori M, Kakizoe T: Pancreatic cancer:
Medical aspects. Pancreas 16:349-354, 1998
62. Saito Y: Report of the nationwide survey of the pancreatic cancer in Japan (1996). [Japanese] Journal of the Japan
Pancreas Society 13:63-91, 1998
63. Semelka RC, Shoenut JR Kroeker MA, et al: Chronic
pancreatitis: MR imaging features before and after administration of gadopentetate dimeglumine. J Magn Reson Imaging
3:79-82, 1993
64. Semelka RC, Ascher SM: MR imaging of the pancreas.
Radiology 188:593-602, 1993
65. Ichikawa T, Haradome H, Hachiya J, et al: Pancreaticductal adenocarcinoma: Preoperative assessment with helical CT
versus dynamic MR imaging. Radiology 202:655-662, 1997
66. Kobayashi G, Fujita N, Lee S, et al: Correlation between
ultrasonographic findings and pathological diagnosis of mucinproducing tumor of the pancreas. [Japanese] Nippon Shokakibyo Gakkai Zasshi 87:235-242, 1990
67. Kasugai K, Hayakawa T, Miyaji M, et al: Mucinproducing pancreatic cancer with unfused main pancreatic
duct--A case report. [Japanese] Nippon Shokakibyo Gakkai
Zasshi 84:2413-2417, 1987
68. Chen J, Baithun SI: Morphological study of 391 cases of
exocrine pancreatic tumours with special reference to the
classification of exocrine pancreatic carcinoma. J Patho1146:1729, 1985
69. Itoh S, Ishiguchi T, Ishigald T, et al: Mucin-producing
pancreatic tumor: CT findings and histopatbologic correlation.
Radiology 183:81-86, 1992
70. Ohhashi K, Murakami Y, Maruyama M, et al: Four cases
of "mucin-producing" cancer of the pancreas. [Japanese] Prog
Dig Endoscopy 20:348-351, 1982
71. Nakagohri T, Asano T, Takayama W, et al: Resection of
the inferior head of the pancreas: Report of a case. Surg Today
26:640-644, 1996
72. Obara T, Maguchi H, Saitoh Y, et al: Mucin-producing
tumor of the pancreas: Natural history and serial pancreatogram
changes. Am J Gastroenterol 88:564-569, 1993
73. Itai Y, Kokubo T, Atomi Y, et al: Mucin-hypersecreting
carcinoma of the pancreas. Radiology 165:51-55, 1987
74. Koito K, Namieno T, Ichimura T, et al: Mucin-producing
pancreatic tumors: Comparison of MR cholangiopancreatography with endoscopic retrograde cholangiopancreatography. Radiology 208:231-237, 1998
75. Sugiyama M, Atomi Y, Hachiya J: Intraductal papillary
tumors of the pancreas: Evaluation with magnetic resonance

MR PANCREATOGRAPHY

cholangiopancreatography [see comments]. Am J Gastroenterol

93:156-159, 1998
76. Itai Y, Ohtomo K, Kokubo T, et al: CT demonstration of
gas in dilated pancreatic duct. J Comput Assist Tomogr 10:10521053, 1986
77. Ammann RW, Muench R, Otto R, et al: Evolution and
regression of pancreatic calcification in chronic pancreatitis. A
prospective long-term study of 107 patients. Gastroenterology
95:1018-1028, 1988

339

78. Rawat B, Fache JS, Burhenne HJ: Extracorporeal shockwave lithotripsy of pancreatic duct stones. Gastrointest Radiol
17:145-147, 1992
79. Bilbao MK, Dotter CT, Lee TG, et al: Complications of
endoscopic retrograde cholangiopancreatography (ERCP). A
study of 10,000 cases. Gastroenterology 70:314-320, 1976
80. Fan ST, Lai EC, Mok FP, et al: Early treatment of acute
biliary pancreatitis by endoscopic papillotomy. N Engl J Med
328:228-232, 1993