| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, The Ohio State University Medical Center, 171 Means
Hall, 1654 Upham Dr., Columbus, OH 43210-1250.
2 Alexander S. Onasis Public Benefit Foundation, Athens, Greece.
3 Ohio Gastroenterology Group, Inc., Columbus, OH.
4 Department of Digestive Diseases, The Ohio State University Medical Center,
Columbus, OH 43230.
Received September 24, 2001;
accepted after revision January 29, 2002.
Presented at the annual meeting of the Radiological Society of North
America, Chicago, November 2000.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. For this retrospective study involving 26 patients with primary sclerosing cholangitis, 31 MR cholangiopancreatograms were compared with 30 endoscopic retrograde cholangiopancreatograms. The MR cholangiopancreatograms were independently interpreted by two abdominal radiologists in a blinded, randomized manner for overall image quality, extent of ductal visualization, and the presence and location of bile duct strictures. Unweighted multirater kappa coefficient values were estimated for each comparison.
RESULTS. Visualization of more than 50% of the expected ductal length was possible in the extrahepatic, central intrahepatic, and peripheral intrahepatic bile ducts in 99%, 88%, and 69% of the MR cholangiopancreatograms and 100%, 86%, and 52% of the endoscopic retrograde cholangiopancreatograms, respectively. Strictures were detected in the extrahepatic, central, and peripheral ducts in 53%, 68%, and 87% of the MR cholangiopancreatograms and 73%, 67%, and 63% of the endoscopic retrograde cholangiopancreatograms, respectively. The interobserver agreement for stricture detection was 61% for MR cholangiopancreatography and 76% for ERCP. MR cholangiopancreatographic findings were consistent with ERCP findings for the presence of strictures in 69% of the cases.
CONCLUSION. In patients with primary sclerosing cholangitis, MR cholangiopancreatography better shows the bile ducts and can depict more strictures, especially of the peripheral intrahepatic ducts, than ERCP. MR cholangiopancreatography can be used to noninvasively diagnose and follow up patients with primary sclerosing cholangitis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Endoscopic retrograde cholangiopancreatography (ERCP) is currently the gold standard for diagnosing primary sclerosing cholangitis [1,2,3]. On ERCP, primary sclerosing cholangitis appears as multifocal areas of stricturing of the intrahepatic and extrahepatic bile ducts with intervening segments of normal or dilated ducts. Although ERCP provides high-quality images, this procedure is invasive and can produce uncommon complications such as sepsis, hemorrhage, pancreatitis, bowel perforation, and cholangitis [4, 5]. Furthermore, ERCP complications occur at a higher rate in patients with primary sclerosing cholangitis than in patients without primary sclerosing cholangitis [6].
MR cholangiopancreatography is a noninvasive and less expensive imaging test than ERCP and has been shown to be as accurate as ERCP for evaluating abnormalities of the extrahepatic bile duct (i.e., stones, strictures) [7,8,9,10]. In contrast to the abundant literature evaluating MR cholangiopancreatography in the characterization of extrahepatic bile duct abnormalities, few studies evaluate MR cholangiopancreatography in the characterization of intrahepatic bile duct abnormalities [10,11,12]. These studies were performed on small groups of patients at single institutions and indicate good correlation between MR cholangiopancreatography and ERCP for the diagnosis of primary sclerosing cholangitis. Furthermore, a recent study by Vitellas et al. [13], comparing MR cholangiopancreatography with ERCP in patients with sclerosing cholangitis, showed that MR cholangiopancreatography may be superior to ERCP with respect to the degree of ductal visualization and rate of stricture detection. However, the differences were not statistically significant.
The purpose of our study was to compare MR cholangiopancreatography with ERCP in patients with primary sclerosing cholangitis for the degree of ductal visualization, the degree of interobserver agreement in stricture detection, and the degree of intermodality agreement for stricture detection. In addition, we compared our results with the results of Vitellas et al. [13] to determine validity.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The study group of 26 patients was composed of 14 men and 12 women with a mean age of 47 years (range, 21-69 years). The duration of primary sclerosing cholangitis ranged from 2 months to 22 years (mean, 4.1 years). The time interval between MR cholangiopancreatography and ERCP ranged from 0 days to 12 months (mean, 4.7 months).
All MR cholangiopancreatograms were acquired using a 1.5-T unit (Signa; General Electric Medical Systems, Milwaukee, WI) with a torso phased array coil and applying the single-shot fast spin-echo pulse sequence. Both thick-slab (20 mm) and thin-slab (5 mm) MR cholangiopancreatograms were acquired during one breath-hold. The thick-slab MR cholangiopancreatograms were obtained in the coronal and coronal oblique planes. Using the common bile duct in the axial scout image as the reference point, we obtained from nine to 12 coronal oblique images at 15° intervals [14]. Thin-slice MR cholangiopancreatograms were obtained in the axial and coronal planes. Imaging parameters included a flip angle of 90°, an infinite TR, a TE that ranged from 900 to 1000 msec, a matrix of 256 x 256 (frequency x phase), and a field of view of 32 cm. In addition, patients were not given glucagon and were not expected to fast before MR cholangiopancreatography. Neither oral nor IV contrast agents were administered.
Endoscopic retrograde cholangiopancreatograms were obtained by one of two gastroenterologists in a standard fashion. Fluoroscopic monitoring and image interpretation were performed by radiologists.
All MR cholangiopancreatograms and endoscopic retrograde cholangiopancreatograms were interpreted independently in a blinded and randomized manner by two abdominal radiologists experienced in cholangiographic interpretation. The observers were unaware of the patients' symptoms, laboratory results, pathologic findings, and ERCP results. The two observers evaluated the overall image quality, extent of ductal visualization, and the presence and location of strictures. Before interpreting the images for this study, the observers reviewed MR cholangiopancreatograms with proven diagnoses of both normal and abnormal findings as an in-service training session to standardize their interpretations of the MR cholangiopancreatograms.
A stricture was defined as a narrowing of a bile duct with proximal dilatation or signal void in the expected location of a bile duct with proximal dilatation. In addition, the bile ducts were categorized into five ductal units: extrahepatic, central right, central left, peripheral right, and peripheral left intrahepatic bile ducts. The endoscopic retrograde cholangiopancreatograms were reviewed independently of the MR cholangiopancreatograms. In addition, the thin- and thick-slab images were interpreted together. Thus, each observer interpreted 31 MR cholangiopancreatograms and 30 endoscopic retrograde cholangiopancreatograms of the 26 patients. In one patient, two MR cholangiopancreatograms were obtained within 1 year of an endoscopic retrograde cholangiopancreatogram.
The overall image quality was rated as very good (no or minimal artifact), fair (degraded but still diagnostic), or poor (nondiagnostic). Each ductal segment (extrahepatic, central, and peripheral) on each cholangiogram was evaluated for the extent of ductal visualization. The extent of ductal visualization was classified as not seen, less than 50%, greater than 50%, or 100% visualization of the expected ductal length.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ductal Visualization
On MR cholangiopancreatography, greater than 50% visualization of the
expected ductal length was noted in the extrahepatic, central, and peripheral
ducts in 99%, 88%, and 69% of the MR cholangiopancreatograms and 100%, 86%,
and 52% of the endoscopic retrograde cholangiopancreatograms, respectively.
Greater than 50% visualization of the expected ductal length was noted in the
central right and left ducts in 86% and 88% of the MR cholangiopancreatograms
and in 77% and 87% of the endoscopic retrograde cholangiopancreatograms,
respectively. Greater than 50% visualization of the expected ductal length was
noted in the peripheral right and left ducts in 61% and 77% of the MR
cholangiopancreatograms and 54% and 50% of the endoscopic
cholangiopancreatograms, respectively. In all, good visualization (>50%
visualization of the expected ductal length) was shown in 82% of the ducts on
MR cholangiopancreatography and in 75% of the ducts on ERCP
(Table 1).
|
Number and Location of Strictures
The total number of ductal units with a stricture detected on MR
cholangiopancreatography and ERCP was 225 and 201 ductal units, respectively.
Strictures of the extrahepatic, central, and peripheral ductal units were
apparent on 53% (n = 33), 68% (n = 84), and 87% (n
= 108) of the MR cholangiopancreatograms and 73% (n = 44), 67%
(n = 81), and 63% (n = 76) of the endoscopic retrograde
cholangiopancreatograms, respectively
(Table 2).
|
Interobserver Agreement for the Presence of Strictures
In total, 155 ductal units were evaluated on 31 MR cholangiopancreatograms
and 150 ductal units were evaluated on 30 endoscopic retrograde
cholangiopancreatograms. The observers agreed on the location of the stricture
in 61% of the MR cholangiopancreatograms (range, 41-72%) and in 76% of the
endoscopic retrograde cholangiopancreatograms (range, 54-89%). The observers
agreed on the presence of strictures of the extrahepatic and intrahepatic
ductal units in 45% and 81% of the MR cholangiopancreatograms and in 86% and
90% of the endoscopic retrograde cholangiopancreatograms, respectively
(Table 3). The percentage of
agreement for the presence of strictures of the central and peripheral
intrahepatic bile ducts was 71% and 74% of the MR cholangiopancreatograms and
73% and 90% of the endoscopic retrograde cholangiopancreatograms,
respectively.
|
Intermodality Agreement for the Presence of Strictures
For the first observer, the percentage of agreement between MR
cholangiopancreatography and ERCP for strictures of the extrahepatic, central
right, central left, peripheral right, and peripheral left ductal units was
50%, 70%, 70%, 63%, and 63%, respectively. The average agreement between MR
cholangiopancreatography and ERCP interpretations for the first observer was
63% (range, 50-70%). For the second observer, the percentage of agreement
between MR cholangiopancreatography and ERCP for stricture location of the
extrahepatic, central right, central left, peripheral right, and peripheral
left ductal units were 82%, 67%, 78%, 81%, and 85%, respectively. The average
agreement between MR cholangiopancreatographic and ERCP interpretations for
the second observer was 79% (range, 67-85%).
When we combined the interpretations of both observers, the interpretations based on MR cholangiopancreatography agreed with those based on ERCP for strictures of the extrahepatic, central, and peripheral intrahepatic bile ducts in 65%, 71%, and 71% of the cases, respectively; the overall percentage agreement was 71% (Table 4). When including only those ducts with greater than 50% visualization of the expected ductal length, the two modalities agreed on the presence of strictures of the extrahepatic, central, and peripheral intrahepatic bile ducts in 66%, 75%, and 91% of the cases, respectively, for an overall agreement of 79%.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Image Quality
Image quality was similar for ERCP and MR cholangiopancreatography.
Ductal Visualization
Compared with ERCP, MR cholangiopancreatography allowed better overall
visualization of the ducts; our findings are similar to the results reported
by Vitellas et al. [13]. As we
expected, the degree of visualization was greatest for the extrahepatic bile
duct and least for the peripheral intrahepatic bile ducts for both imaging
modalities. Both modalities showed the extrahepatic and central ducts to a
similar extent. However, MR cholangiopancreatography was superior to ERCP in
showing the peripheral ducts (69% vs 52%, respectively); these findings are
similar to those reported by Vitellas et al.
[13]. Unlike ERCP. which
requires contrast material for opacification of the bile ducts, MR
cholangiopancreatography images the bile ducts in a physiologic nondistended
state and relies on the internal contrast of bile on T2-weighted sequences for
the illustration of the bile ducts. Therefore, strictures, thick bile, or
stones in patients with primary sclerosing cholangitis—all of which may
impede contrast opacification and limit the visualization of the proximal
ducts on ERCP—do not limit the visualization of these ducts on MR
cholangiopancreatography. In fact, these abnormalities might produce
dilatation of the ducts and therefore allow better visualization of the ducts
on MR cholangiopancreatography (Figs.
1A,1B
and
2A,2B).
|
|
|
|
This study confirms results from prior studies that have shown that the degree of visualization of the peripheral intrahepatic duct on MR cholangiopancreatography in patients with sclerosing cholangitis differs from that in healthy volunteers and control subjects [12, 13]. Good visualization of the peripheral intrahepatic bile ducts on MR cholangiopancreatography should alert the radiologist to the presence of strictures.
In this study, we showed that the extrahepatic bile duct can be visualized to a similar degree on MR cholangiopancreatography and ERCP. However, we did not categorize the degree of ductal visualization above 90% or 100% visualization of the expected ductal length as was done by Fulcher et al. [12]. If we had categorized the degree of ductal visualization as greater than 90% and 100% visualization of the expected ductal length, our results might have shown that the extrahepatic bile duct was better visualized on ERCP (Figs. 3A,3B and 4A,4B).
|
|
|
|
Stricture Visualization
The overall rate of stricture detection was similar with MR
cholangiopancreatography and ERCP. The prior study by Vitellas et al.
[13] showed that MR
cholangiopancreatography was superior to ERCP in stricture detection, but the
difference was not statistically significant.
As expected in patients with sclerosing cholangitis, peripheral intrahepatic bile duct strictures predominated on both imaging modalities. MR cholangiopancreatography, which offered better visualization of the peripheral bile ducts, also showed more peripheral duct strictures than ERCP. These results are similar to those reported by Vitellas et al. [13] (Figs. 1A,1B and 2A,2B).
Strictures of the extrahepatic bile duct were more often visualized on ERCP (Figs. 3A,3B and 4A,4B) than on MR cholangiopancreatography. On ERCP, the infusion of contrast material via a catheter in the distal common bile duct produces distention of the duct that potentially enables better visualization of the duct [16]. The distention of the extrahepatic bile duct and its proximity to the catheter probably account for the increase in stricture detection on ERCP. On the other hand, proximal strictures of the extrahepatic bile duct potentially can cause a decrease in the volume of bile in the duct distal to the stricture, resulting in less distention and decreased visualization. The decreased visualization of the extrahepatic bile duct in the setting of proximal strictures probably explains why strictures of the extrahepatic bile duct are detected less often on MR cholangiopancreatography than on ERCP. This theory is supported by Fulcher et al. [12], who showed a statistically significant difference (p = 0.001) between complete delineation of the extrahepatic bile duct on MR cholangiopancreatography in patients with sclerosing cholangitis (74%) and control subjects (100%). Therefore, incomplete visualization of the extrahepatic bile duct on MR cholangiopancreatography in patients with sclerosing cholangitis should suggest a stricture.
Strictures of the central intrahepatic bile duct were seen with similar frequency on ERCP and MR cholangiopancreatography.
Interobserver Agreement
The observers agreed more often about the presence of strictures on ERCP
for all ductal units. On MR cholangiopancreatography, the percentage of
agreement for stricture detection progressively increased from the
extrahepatic to the peripheral intrahepatic bile ducts (45-74%). In fact,
interobserver agreement increased by 80% when comparing agreement for
strictures of the extrahepatic and intrahepatic bile ducts on MR
cholangiopancreatography. This increase in agreement probably is associated
with the progressive increase in the degree of ductal visualization when
comparing extrahepatic and central intrahepatic bile ducts with the peripheral
intrahepatic bile ducts on MR cholangiopancreatography.
Although the percentage of agreement increased from the extrahepatic to the
peripheral intrahepatic bile ducts on ERCP (86-90%), the increase was not
progressive or as great as that on MR cholangiopancreatography. Our results
support the findings of Fulcher et al.
[12], who showed that the
interobserver agreement for the diagnosis of sclerosing cholangitis on ERCP
(
= 1.0) was greater than on MR cholangiopancreatography ( =
0.79). In addition, their study showed that the interobserver agreement for
intrahepatic sclerosing cholangitis ( = 0.71) was greater than the
interobserver agreement for extrahepatic sclerosing cholangitis ( =
0.62) on MR cholangiopancreatography.
In patients with sclerosing cholangitis, therapy is not based on the number of strictured ducts, but on the presence of dominant strictures amendable to percutaneous or endoscopic therapy. Thus, even though the number of strictured ducts detected by each observer varied, the clinical treatment of these patients should not be altered because, in our experience, MR cholangiopancreatography is accurate in showing dominant strictures in patients with sclerosing cholangitis.
The discrepancies regarding the number of strictured ducts detected by each observer may have more clinical significance in terms of the diagnosis of sclerosing cholangitis. In our study, strictures of the intrahepatic and extrahepatic bile ducts that were consistent with sclerosing cholangitis were depicted in 92% and 53% of the MR cholangiopancreatograms and 97% and 73% of the endoscopic retrograde cholangiopancreatograms, respectively. However, the observers in this study were aware of clinical history before they reviewed the cholangiograms. Previous studies in which observers were unaware of clinical history before reviewing cholangiograms showed that MR cholangiopancreatography is accurate in the diagnosis of sclerosing cholangitis [12, 13]. However, false-negative MR cholangiographic results may occur in patients with cirrhosis or disease limited to the intrahepatic bile ducts [12] (Fig. 5A,5B).
|
|
Intermodality Agreement
In this study, we showed that the overall agreement between MR
cholangiopancreatography and ERCP was greatest for detecting strictures of the
intrahepatic bile ducts (Figs.
4A,4B,
6A,6B,
and
7A,7B).
However, because the degree of ductal visualization differs between the two
modalities, an accurate comparison may not be possible if all ductal segments
are included in the analysis, regardless of the degree of ductal
visualization. When we evaluated only ducts with greater than 50%
visualization of the expected ductal length, the percentage of agreement for
detection of strictures in the extrahepatic and intrahepatic bile ducts
increased to 66% and 83%, respectively. The greatest change occurred in the
percentage of agreement of stricture detection in the peripheral intrahepatic
bile ducts (71-91%) and the least change occurred in the percentage of
agreement of stricture detection in the extrahepatic bile duct (65-66%). Our
results are similar to those reported by Vitellas et al.
[13]. Interestingly, in that
study, most of the disagreements occurred because a stricture detected on MR
cholangiopancreatography was not identified on ERCP. This finding suggests
that either MR cholangiopancreatography is more sensitive for the detection of
strictures or MR cholangiopancreatography produces false-positive findings
that mimic strictures. A multiinstitutional study comparing ducts that are
completely visualized on both imaging modalities in a large group of patients
with sclerosing cholangitis is needed to enable a more accurate comparison
between MR cholangiopancreatography and the gold standard ERCP.
|
|
|
|
In this study, we categorized the bile ducts into five ductal units. Separating the bile ducts into small units should provide more accurate results about the level of agreement between MR cholangiopancreatography and ERCP. However, this method of categorization can also increase the number of discordant interpretations. Results from prior studies that separated the bile ducts into fewer ductal units have shown higher intermodality agreement [11, 12].
Our study has several limitations. First, because the degree of ductal visualization differs for each imaging modality, accurately comparing the two modalities for the detection of strictures may not be possible. A study comparing ducts that are completely visualized on both imaging modalities would provide more accurate comparisons. Second, even though the average interval between MR cholangiopancreatography and ERCP was 4.7 months, the interval in several patients approached 12 months. Longer intervals between cholangiographic studies could have allowed the disease to progress and produce more strictures. Third, several variables affect the degree of ductal opacification and visualization of the bile ducts on ERCP. These variables include the experience of the endoscopist, catheter position, infusion pressure, and cooperation of the patient. Fourth, the duration of primary sclerosing cholangitis averaged 4.1 years. Only a few patients had advanced disease with cirrhosis. A similar study using a population with more advanced disease may show less visualization and therefore fewer strictures of the intrahepatic bile ducts. Fifth, the observers were aware of each patient's clinical history before reviewing the cholangiograms. Sixth, the level of experience of the two observers (3 years vs 12 years) differed considerably and could partially account for the degree of interobserver variability.
In conclusion, the results of this study validate those described by Vitellas et al. [13]. In patients with primary sclerosing cholangitis, MR cholangiopancreatography allows better visualization of and depicts more strictures of the peripheral intrahepatic bile ducts than ERCP. Therefore, our study provides additional support for the use of noninvasive MR cholangiopancreatography for the diagnosis and follow-up of patients with primary sclerosing cholangitis.
Acknowledgments
The American College of Radiology's Technology Assessment Studies
Assistance Program assisted this research by providing the professional
services of Philip E. Crewson, Craig A. Beam, and Rebecca Lewis in addition to
extensive data entry.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  V Shanmugam, G C Beattie, S R Yule, W Reid, and M A Loudon Is magnetic resonance cholangiopancreatography the new gold standard in biliary imaging? Br. J. Radiol., October 1, 2005; 78(934): 888 - 893. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D P O'Regan, J Fitzgerald, J Allsop, D Gibson, D J Larkman, D Cokkinos, J V Hajnal, and S A Schmitz A comparison of MR cholangiopancreatography at 1.5 and 3.0 Tesla Br. J. Radiol., October 1, 2005; 78(934): 894 - 898. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Aube, B. Delorme, T. Yzet, P. Burtin, J. Lebigot, P. Pessaux, C. Gondry-Jouet, J. Boyer, and C. Caron MR Cholangiopancreatography Versus Endoscopic Sonography in Suspected Common Bile Duct Lithiasis: A Prospective, Comparative Study Am. J. Roentgenol., January 1, 2005; 184(1): 55 - 62. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
