| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Department of Radiology (A.S.F., M.A.T.) and the Hepatology Section (K.J.F., M.L.S., R.K.S., V.A.C.L., A.J.S.), Medical College of Virginia of Virginia Commonwealth University, 401 N 12th St, Main Hospital, 3rd Fl, Rm 3-415, Richmond, VA 23298-0615. Received June 18, 1999; revision requested July 30; revision received August 23; accepted August 30. Address reprint requests to A.S.F. (e-mail: asfulche@hsc.vcu.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: In a prospective case-control study involving 102 patients, the MR cholangiograms obtained in 34 patients with PSC established with endoscopic retrograde cholangiopancreatography (ERCP) were compared with the MR cholangiograms obtained in 68 age-matched control patients with hepatobiliary diseases other than PSC. Two abdominal radiologists conducted an independent, blinded random review of the MR cholangiograms to assess for the presence or absence of PSC and determine the location of PSC in the biliary tract, and then compared the findings with those at ERCP.
RESULTS: MR cholangiography was found to be accurate in detecting PSC and in defining the extent of disease. In the detection of PSC, the sensitivities were 88% and 85%; specificities, 97% and 92%; positive predictive values, 94% and 85%; and negative predictive values, 94% and 93% for readers 1 and 2, respectively. Interobserver agreement was excellent (
= 0.79). In the localization of extrahepatic PSC, the sensitivities were 83% and 89%; and specificities, 83% and 83% for readers 1 and 2, respectively. Interobserver agreement was good ( = 0.62). In the localization of intrahepatic PSC, the sensitivity was 87% for both readers; interobserver agreement was good ( = 0.71).
CONCLUSION: MR cholangiography enables accurate detection and localization of PSC.
Index terms: Bile ducts, MR, 76.121415, 76.121419 • Cholangitis, 76.288 • Endoscopic retrograde cholangiopancreatography (ERCP), 76.1222 • Magnetic resonance (MR), half-Fourier imaging, 76.121415, 76.12419
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The diagnosis of PSC is usually established with endoscopic retrograde cholangiopancreatography (ERCP) and less often with percutaneous transhepatic cholangiography; these studies demonstrate multiple segmental strictures, mural irregularities, and diverticula that are characteristic of PSC (6–8). Although ERCP and percutaneous transhepatic cholangiography provide high-quality images of the biliary tract, both are invasive procedures that are associated with risks of sepsis and hemorrhage (9,10). ERCP is also associated with risks of pancreatitis and bowel perforation (11,12) and has been shown to result in progression of cholestasis in patients with advanced PSC (13).
For these reasons, a noninvasive examination that allows high-quality imaging of the biliary tract without the associated risks of ERCP and percutaneous transhepatic cholangiography is needed for the diagnosis and surveillance of patients with PSC. Although ultrasonography (US) and computed tomography (CT) may enable the identification of thickening of the bile duct wall and biliary dilatation (14–16), these modalities are not widely accepted for establishing the diagnosis of PSC. Magnetic resonance (MR) cholangiography might be useful in determining the presence or absence of PSC and in depicting the extent of ductal involvement. In a retrospective study involving nine patients with PSC and 20 healthy volunteers conducted by Ernst et al (17), MR cholangiography provided diagnostic information that was comparable to that provided by ERCP. In this preliminary study, the sensitivity and specificity of MR cholangiography in the diagnosis of PSC was 100%; the high specificity was related in part to the healthy control group.
To establish the sensitivity and specificity of MR cholangiography in the evaluation of PSC in a setting that more closely approximates that in clinical practice, this study was conducted to compare a group of patients with PSC with a group of patients with hepatobiliary disease rather than with a group of healthy control subjects. The primary purpose of this prospective study was to determine the accuracy of MR cholangiography in detecting PSC and in distinguishing patients with PSC from control patients with known hepatobiliary diseases other than PSC. A secondary purpose was to compare the accuracy of MR cholangiography with that of ERCP in determining the location of ductal involvement by PSC.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The patients with PSC and the control patients were recruited by means of a search of the hospital, hepatology, and radiology databases at our institution. The key words used to search for patients with PSC included "primary sclerosing cholangitis" and "sclerosing cholangitis." The key words used for the age-matched control patient search included "cholestatic liver disease," "cholestasis," "cirrhosis," "primary biliary cirrhosis," "hepatitis," "autoimmune hepatitis," "hepatitis B virus," and "hepatitis C virus." The inclusion criterion for patients with PSC was a diagnosis of PSC that was established by using direct cholangiography and supported by the clinical presentation and course and the results of biochemical analysis. The inclusion criterion for the age-matched control patients was an established diagnosis of a hepatobiliary disease other than PSC. The exclusion criteria for both groups included pregnancy, inability or refusal to provide informed consent, and conditions precluding entry into the MR imager, such as cardiac pacemakers, cerebral aneurysm clips, any other foreign bodies incompatible with a magnetic field; severe claustrophobia; and morbid obesity.
The medical charts of the patients with PSC and the control patients were reviewed by an author (K.J.F.) who was uninvolved in image interpretation. A standardized data sheet was completed for each patient. The information obtained from the chart review included age, sex, method by which PSC or other hepatobiliary disease was established, duration of disease, liver chemistry profile, history of inflammatory bowel disease, results of liver biopsy, and results of radiologic examinations, including ERCP, US, CT, and MR imaging. In particular, the presence or absence of cirrhosis noted on previously obtained cross-sectional images was recorded. Cirrhosis was considered to be present at cross-sectional imaging if hypertrophy of the lateral segment of the left hepatic lobe or caudate lobe and/or atrophy of the medial segment of the left hepatic lobe, atrophy of the anterior segment of the right hepatic lobe, or contour lobulation or nodularity was identified (18–21).
Imaging
MR cholangiograms were obtained with a 1.0-T magnet (Magnetom Expert; Siemens Medical Systems, Erlangen, Germany) (maximum gradient strength, 20 mT/m; rise time, 1 msec) programmed with the half-Fourier RARE sequence. A circularly polarized, phased-array body coil was used in all cases. Antiperistaltic agents and oral contrast material were not used, and the patients did not fast prior to the examination. Initial imaging of the abdomen was performed in the coronal plane without fat suppression and with the use of a half-Fourier RARE sequence to obtain scout images of the abdomen (infinite/60 [repetition time msec/effective echo time msec], 150° refocusing flip angle, 8-mm section thickness, 10% intersection gap, 400 x 400-mm field of view, one signal acquired, 208 x 256 [phase encoding x frequency encoding] matrix, 22-second acquisition time). The biliary tract was then localized with fat-suppressed, thick-section RARE images in the coronal-oblique (25°) and transverse planes. The parameters used to obtain the thick-section MR cholangiograms included infinite/1,100 (effective), 180° refocusing flip angle, 40-mm section thickness, 270 x 270-mm field of view, one signal acquired, 240 x 256 (phase encoding x frequency encoding) matrix, and 7-second acquisition time.
The thick-section MR cholangiograms were then used as guides to prescribe the appropriate angles of acquisition for the multisection, thin-section (5-mm) MR cholangiographic technique. These MR cholangiograms were acquired in the coronal-oblique plane parallel to the long axis of the extrahepatic bile duct. MR technologists with a minimum of 2 years experience in performing abdominal MR imaging prescribed the angles of acquisition that delineated the biliary tract. All MR cholangiograms were reviewed at the console by an abdominal radiologist (A.S.F.) before the patient was removed from the magnet. Additional angles of acquisition were obtained to depict the biliary tract if it was deemed necessary by this radiologist. The parameters for the multisection, thin-section technique included infinite/88.0 (effective), 140° refocusing flip angle, 5.0-mm section thickness with no intersection gap, 270 x 270-mm field of view, one signal acquired, 240 x 256 (phase encoding x frequency encoding) matrix, and an 18-second acquisition time. Thirteen images were obtained during each 18–20-second acquisition. Fat saturation was used in all cases. Both the thick-section and thin-section images were obtained during a breath hold. Maximum intensity projection and multiplanar reformatting techniques were applied to the acquired data; postprocessing was performed at a standard console. Conventional MR imaging of the liver with T1- and T2-weighted sequences was not performed.
No identifying patient information appeared on the MR cholangiograms except for a number assigned to each patient upon enrollment into the study. Breast tissue was excluded from the field of view so that the sex of the patient could not be determined from the MR cholangiograms.
ERCP was performed by gastroenterologists; fluoroscopic monitoring and film hard-copy interpretation of these images were conducted by radiologists. The ERCP images were recorded by using analog and digital technology. The interval between the performance of ERCP and MR cholangiography ranged from 0 days to 117 months (mean, 14 months).
Image Analysis
The MR cholangiograms were interpreted independently and prospectively by two experienced abdominal radiologists (A.S.F., M.A.T.) who were blinded to patient identification, patient mix (ie, number of patients with PSC vs number of control patients), and all clinical, laboratory, pathologic, and prior imaging examination findings. The source images as well as the three-dimensional reconstructions were reviewed on film hard copy and at an interactive workstation. However, all diagnostic decisions were made on the basis of the source image findings. The MR cholangiographic findings were recorded on standardized data sheets.
The MR cholangiograms were evaluated for delineation of the first-, second-, third-, and fourth-order intrahepatic bile ducts; the extrahepatic bile duct; and the gallbladder in the patients with PSC and the control patients. The following grading system was used: excellent for complete delineation, good for delineation of at least 90%, fair for delineation of less than 90%, and nondepiction.
An assessment of the presence or absence of PSC at MR cholangiography was made on the basis of standard cholangiographic features, which include multiple segmental strictures of the extrahepatic and intrahepatic bile ducts, alternating strictures and dilatation of the ducts, and mural irregularities (6–8). The certainty of the interpreting radiologists in determining the presence or absence of PSC was recorded with the use of a five-point rating system of definitely present, probably present, equivocal, probably absent, and definitely absent. After the MR cholangiograms had been reviewed by the two interpreting radiologists, their interpretations regarding the presence or absence of PSC were compared with the known disease processes in each patient on the basis of chart review for all the patients and the ERCP image findings (n = 24) and ERCP reports (n = 10) for the patients with PSC. The ERCP images could not be retrieved for 10 patients. However, in each of these 10 patients, the diagnosis of PSC had been established with ERCP and was supported by the results of clinical and biochemical analyses and liver biopsy.
The MR cholangiograms were also evaluated for the location of disease as intrahepatic, extrahepatic, or both in those patients with PSC whose ERCP images were available for interpretation (n = 24). The certainty of the interpreting radiologists in determining the location of PSC was recorded with the use of a five-point rating system of definitely present, probably present, equivocal, probably absent, and definitely absent. After the MR cholangiograms had been interpreted independently by both study radiologists, the same two radiologists reviewed the available ERCP images of the patients with PSC in a random, independent, blinded manner and recorded the location of PSC as intrahepatic, extrahepatic, or both. The determination of disease location made on the basis of the ERCP image interpretations was used as the standard of reference. To avoid memory-recollection bias, there was a minimum interval of 4 weeks between the reviews of the MR cholangiograms and ERCP images. Following the interpretation of the MR cholangiograms and ERCP images, the accuracy of assessment of disease location at MR cholangiography was determined for each of the interpreting radiologists.
An assessment was made regarding the presence of bile duct stones at MR cholangiography. Bile duct stones were diagnosed as well-defined, low-signal-intensity filling defects in the high-signal-intensity bile. ERCP or CT was used to confirm the presence of intraductal stones.
Statistical Analysis
Prior to the initiation of the study, a sample size calculation was conducted to determine the 95% CIs for sensitivity. The Wilcoxon rank sum test was used to compare depictions of the ducts at MR cholangiography for the PSC and control groups. Receiver operating characteristic (ROC) curve analyses were conducted to assess the diagnostic performance of MR cholangiography in the determination of the presence or absence of PSC and in the localization of PSC. The area under the curve Az and the standard error of the Az were calculated by using the parametric trapezoidal rule. The Az values for the two readers were compared.
The sensitivity, specificity, positive predictive value, and negative predictive value of MR cholangiography for the detection and localization of PSC were calculated. For determination of the sensitivity and specificity values, definitely present and probably present ratings were considered to be positive, and definitely absent, probably absent, and equivocal ratings were considered to be negative.
Interobserver agreement for determination of the presence or absence of PSC and for localization of PSC was calculated by using the statistic. The level of agreement was defined as follows: value less than 0.40 for poor agreement, value of 0.40–0.75 for good agreement, and value greater than 0.75 for excellent agreement.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The control group consisted of 68 patients who ranged in age from 25 to 73 years (mean age, 45 years) and had hepatobiliary diseases other than PSC: 24 of these patients had biliary cirrhosis; 12, hepatitis C virus; six, autoimmune hepatitis; three, alcoholic or cryptogenic cirrhosis; 18, idiopathic cholestatic liver disease; three, cirrhosis secondary to Budd-Chiari syndrome; and two, hepatic tumors (one adenocarcinoma and one lymphoma). The number of years for which hepatobiliary disease had been established in the control group ranged from 1 to 15 (mean, 3.4 years). Liver biopsy had been performed in 44 of the 68 patients, and its results revealed cirrhosis in 13 patients. Fifty-four of the 68 control patients had undergone abdominal US, CT, and/or MR imaging. The results of these cross-sectional imaging examinations demonstrated morphologic evidence of cirrhosis in 21 of the 54 patients: in nine with US, in seven with MR imaging and US, in four with MR imaging only, and in one with US and CT.
MR Cholangiographic Depiction of the Biliary Tract
In the PSC group (n = 34), MR cholangiography permitted the complete demonstration of the first-, second-, third- and fourth-order intrahepatic bile ducts in 32 (94%), 33 (97%), 29 (85%), and 18 (53%) patients, respectively (Table 1). All of the 18 patients in whom the fourth-order bile ducts were depicted completely and all of the 29 patients in whom the third-order ducts were depicted completely had evidence of intrahepatic PSC. The intrahepatic ductal strictures and the associated ductal dilatation proximal to the strictures permitted visualization of the more peripherally located ducts. In contrast, in the control group (n = 68), the third- and fourth-order bile ducts were depicted completely in 39 (57%) and 22 (32%) patients, respectively (Table 2). The difference in complete depiction of the third-order bile ducts between the PSC and control groups was statistically significant (P = .007). In three of the five patients with PSC in whom the first-, second-, or third-order bile ducts were not depicted at MR cholangiography, extensive cirrhosis was present at conventional MR imaging (n = 2) and US (n = 1), and this presumably accounted for the inability to visualize the intrahepatic bile ducts.
|
|
The number of thin-section acquisitions required to depict the biliary tract ranged from three to 10 (mean, four) in the PSC group and from three to six (mean, four) in the control group. The room time required for acquisition of the images ranged from 7 to 12 minutes (mean, 10 minutes) for the patients with PSC and from 7 to 9 minutes (mean, 8 minutes) for the control group.
Determination of the Presence or Absence of PSC
MR cholangiography permitted the identification of intrahepatic and extrahepatic multifocal strictures, mural irregularities, and ductal dilatation characteristic of PSC (Figs 1–3). ROC curve analysis demonstrated Az values of 0.9734 and 0.9414 for readers 1 and 2, respectively (Fig 4). There was no significant difference in the ROC curves between the two readers (P = .28). The sensitivity, specificity, positive predictive value, and negative predictive value of MR cholangiography in the evaluation for PSC were 88%, 97%, 94%, and 94%, respectively, for reader 1 and 85%, 92%, 85%, and 93%, respectively, for reader 2 (Table 3). Interobserver agreement in the determination of the presence or absence of PSC was excellent ( = 0.79).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Determination of the Location of PSC
Of the 34 patients with an established diagnosis of PSC, 24 had ERCP images that could be retrieved for review. The independent, random review of these cholangiograms, conducted by the two radiologists who had previously reviewed the MR cholangiograms, revealed extrahepatic PSC in 18 of 24 patients and intrahepatic PSC in all 24 patients. The two radiologists agreed on the interpretation of the ERCP images with regard to the location of PSC in all instances ( = 1.0). The location of PSC noted at ERCP was considered to be the standard of reference and was compared with the MR cholangiographic determination of the location of PSC by each of the two readers.
ROC curve analysis demonstrated Az values of 0.8667 and 0.8848 for readers 1 and 2, respectively, in the identification of extrahepatic PSC (Fig 8). There was no statistically significant difference in the ROC curves between the two readers (P = .89). The sensitivity, specificity, positive predictive value, and negative predictive value for the identification of extrahepatic bile duct PSC were 83%, 83%, 94%, and 63% for reader 1 and 89%, 83%, 94%, and 71% for reader 2 (Table 3); interobserver agreement was good ( = 0.62). In two patients, both readers incorrectly assessed the presence or absence of extrahepatic PSC. In one of these two patients, the readers did not detect changes of PSC. In this patient, the ERCP image showed extrahepatic PSC, but the extrahepatic bile duct was incompletely depicted at MR cholangiography owing to cirrhosis that resulted in its reorientation. In the second patient, both readers noted an irregularity of the extrahepatic bile duct at MR cholangiography that could not be confirmed at ERCP.
|
= 0.71) (Table 3). Because all 24 patients had evidence of intrahepatic PSC at ERCP, ROC curve analysis could not be conducted. Twenty-one patients were correctly classified as having intrahepatic PSC by both readers. Two patients with intrahepatic PSC were classified as having no MR cholangiographic evidence of intrahepatic PSC by both readers. In one of these two patients, the intrahepatic ducts were not well depicted because of cirrhosis. In the second patient, PSC was not identified at MR cholangiography owing to its predominance in the third- and fourth-order bile ducts (Fig 6). Each of the readers incorrectly classified one additional patient as having no evidence of PSC involving the intrahepatic bile ducts. In two patients, the abnormal segments of the intrahepatic bile ducts were demonstrated more completely at MR cholangiography than at ERCP. In one patient, the abnormal ductal segment was filled with stones (Fig 9), and in the second patient, a branch of the left hepatic duct was involved by a high-grade stricture that prevented complete opacification at ERCP.
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In addition to complications, ERCP is associated with failed cannulation of the bile or pancreatic duct in 10%–20% of patients (9). In contrast, half-Fourier RARE MR cholangiography has been shown to produce diagnostic images of the pancreaticobiliary tract in more than 99% of patients (22,23). In the present study, the first-order intrahepatic bile ducts were depicted completely in 32 (94%) of 34 patients with PSC and in 66 (97%) of 68 control patients. The second-order intrahepatic bile ducts were depicted completely in 33 (97%) of 34 patients with PSC and in 64 (94%) of 68 control patients. However, despite the high rate of demonstration of the first- and second-order bile ducts, MR cholangiography depicted the third- and fourth-order ducts less frequently, especially in nondilated systems, in part because of the lower spatial resolution of MR cholangiography compared with that of ERCP. Technical developments in MR hardware and software may result in improved spatial resolution on MR cholangiograms.
At least partial depiction of the extrahepatic bile duct was achieved in all the patients with PSC, and complete depiction was achieved in all the control patients. Although the thin-section (5-mm) MR cholangiograms were useful in depicting focal strictures and subtle irregularities of ductal segments on multiple images, the thick-section (40-mm) MR cholangiograms provided a comprehensive view of the entire biliary tract on a single image (Fig 1a–1c). For these reasons, we recommend the use of both thin- and thick-section MR cholangiopancreatographic techniques in the evaluation of patients with PSC.
Although good depiction of the biliary tract is achieved with ERCP, particularly when it is performed with balloon occlusion techniques, segments of the intrahepatic bile ducts may not be delineated at ERCP owing to high-grade strictures or impacted intraductal stones. In contrast, MR cholangiography allows visualization of ducts both proximal and distal to high-grade strictures, because depiction of the ducts with MR cholangiography does not depend on achieving high injection pressures. In the setting of impacted stones in a ductal segment, only a small amount of bile must to be present to achieve depiction of the duct and stones at MR cholangiography.
For MR cholangiography to serve as a feasible alternative to diagnostic ERCP for the detection of PSC, MR cholangiography must be able to not only depict the bile ducts but also delineate the multifocal strictures, ductal dilatation, and mural irregularities that are characteristic of PSC. In the present study, the sensitivity and specificity of MR cholangiography for the detection of PSC were 88% and 97%, respectively, for one reader and 85% and 92%, respectively, for the second reader; interobserver agreement was excellent.
The difficulties in interpreting the MR cholangiograms led to false-positive and false-negative diagnoses and were related to two factors: the presence of cirrhosis and PSC limited to the peripheral intrahepatic bile ducts. Because each of the false-positive diagnoses of PSC occurred in patients with cirrhosis, it is possible that the utility of MR cholangiography in excluding PSC may be limited in this setting. Although cirrhosis may result in distortion of the intrahepatic bile ducts and lead to false-positive diagnoses of PSC, conversely, the altered hepatic architecture associated with cirrhosis may make visualization of abnormalities of the intrahepatic and extrahepatic bile ducts more difficult and thus lead to false-negative diagnoses of PSC. In particular, the hypertrophy and atrophy of various hepatic segments that are associated with cirrhosis may result in reorientation of the extrahepatic bile duct, which, in turn, makes visualization of the duct and ductal abnormalities problematic.
Two of the six false-negative diagnoses in the present study occurred in patients with cirrhosis. In addition to cirrhosis, PSC limited to the peripheral intrahepatic ducts may be difficult to detect at MR cholangiography and thus result in false-negative diagnoses. In the present study, PSC limited to the intrahepatic bile ducts, especially the third- and fourth-order ducts, accounted for three of the false-negative diagnoses.
Although ERCP was used as the standard of reference in this study, this procedure may also result in false-positive and false-negative diagnoses of PSC. False-positive diagnoses may occur because of incomplete biliary tract distention that mimics the ductal irregularities of PSC. In contrast to ERCP, MR cholangiography offers the advantage of depicting the biliary tract in its natural degree of distention. Terada and Nakanuma (24) noted that a PSC-like cholangiographic appearance of the bile ducts may be seen in association with diseases such as cirrhosis and hepatocellular carcinoma and lead to false-positive diagnoses of PSC at direct cholangiography. False-negative diagnoses at ERCP may be related to inadequate opacification of the intrahepatic ducts. This inadequate ductal opacification may be due to pathologic processes of the ducts such as high-grade strictures. Alternatively, incomplete ductal opacification may be technical in nature and result in the failure to detect abnormalities such as strictures.
In addition to determining the presence or absence of PSC, MR cholangiography may be used to assess the location of the disease within the biliary tract. In the present study, although both readers correctly localized disease within the biliary tract and achieved good interobserver agreement, the determination of the presence or absence of intrahepatic disease was negatively affected by the presence of cirrhosis and disease limited to the peripheral intrahepatic ducts. A major advantage afforded by the accurate localization of disease with MR cholangiography is the ability to identify subsets of patients who might benefit from therapeutic ERCP, such as those with dominant extrahepatic bile duct strictures (25).
In the current study, MR cholangiography was deliberately analyzed in the absence of conventional MR imaging to determine the ability of MR cholangiography alone to enable assessment for findings of PSC. It is likely that MR cholangiography performed in conjunction with conventional MR imaging of the abdomen would result in improved accuracy in the diagnosis of PSC. Recently, Dodd et al (20) and Ito et al (21) at CT and MR imaging, respectively, identified alterations in the liver morphology that occur more frequently in patients with PSC than in those with other causes of cirrhosis. These alterations include caudate lobe hypertrophy and atrophy of the lateral segment of the left hepatic lobe. Additional supportive evidence of PSC includes lymphadenopathy and bile duct calculi (26,27). Unlike ERCP, MR permits imaging of the solid organs and lymph nodes and provides additional diagnostic information when it is performed with MR cholangiography.
In the present study, none of the patients in the PSC study group had pathologically proved cholangiocarcinoma. The accuracy of MR cholangiography in detecting cholangiocarcinoma in the setting of PSC has not been determined. In a preliminary study, Fulcher and Turner (28) reported on the utility of MR cholangiography in the detection and staging of hilar cholangiocarcinoma in patients without PSC. However, in patients with PSC, the MR cholangiographic differentiation between a benign ductal stricture of PSC and a malignant stricture due to cholangiocarcinoma is probably more problematic, as is the case with direct cholangiography. In an early report on the direct cholangiographic appearances of cholangiocarcinoma complicating PSC, MacCarty et al (29) noted the lack of specificity of cholangiography in distinguishing a benign stricture from a stricture due to cholangiocarcinoma. In a more recent study conducted by Campbell et al (30), the results confirmed the nonspecificity of a dominant stricture at direct cholangiography, and the utility of cross-sectional imaging studies such as MR in demonstrating cholangiocarcinomas was noted.
Although MR cholangiography offers several advantages compared with ERCP, the major disadvantage of MR cholangiography is that it is a purely diagnostic examination. Unlike ERCP, MR cholangiography does not provide access for interventional procedures such as stent placement and brush biopsy.
Several limitations of this study exist. One limitation is that the interval between the performance of ERCP and MR cholangiography was lengthy for some patients. However, the relative stability of the cholangiographic findings of PSC, as noted by MacCarty et al (29), may have minimized the importance of this limitation. A second limitation is that even though the disease in all the patients with PSC was documented at ERCP and supported by the clinical course, biochemical analysis, and/or pathologic analysis, the ERCP images obtained in all these patients were not available for review. A third limitation is that the mean duration of disease in the patients with PSC was 7.5 years, which may correlate with advanced disease. Additional series focusing on patients with recently diagnosed PSC are required to determine the accuracy of MR cholangiography in detecting early-stage PSC. However, the value of early diagnosis of PSC may be limited because currently there is no effective treatment.
In conclusion, the data in this study supporting the use of MR cholangiography in the detection and localization of PSC are preliminary and require validation with large multicenter trials. Nevertheless, MR cholangiography provides a noninvasive alternative to diagnostic ERCP for the detection of PSC and may assist in identifying patients who might benefit from therapeutic ERCP. Our experience in this study demonstrates the ability of MR cholangiography to provide images of the biliary tract that are of high quality and can be used to detect and localize the cholangiographic findings of PSC and accurately distinguish patients with PSC from those with other hepatobiliary diseases.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, A.S.F.; study concepts, all authors; study design, A.S.F., M.A.T., K.J.F., M.L.S.; definition of intellectual content, A.S.F., M.A.T., K.J.F., M.L.S.; literature research, A.S.F.; clinical studies, A.S.F., K.J.F., A.J.S., V.A.C.L., R.K.S.; data acquisition, A.S.F., K.J.F., A.J.S., V.A.C.L., R.K.S.; data analysis, A.S.F., M.A.T.; manuscript preparation, A.S.F.; manuscript editing, A.S.F., M.A.T.; manuscript review, A.S.F., M.A.T., K.J.F., M.L.S.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Novellas, T. Caramella, M. Fournol, J. Gugenheim, and P. Chevallier MR Cholangiopancreatography Features of the Biliary Tree After Liver Transplantation Am. J. Roentgenol., July 1, 2008; 191(1): 221 - 227. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Q. Knowlton, A. J. Taylor, M. Reichelderfer, and J. Stang Imaging of Biliary Tract Inflammation: An Update Am. J. Roentgenol., April 1, 2008; 190(4): 984 - 992. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Alvaro, G. Macarri, M. G. Mancino, M. Marzioni, M. Bragazzi, P. Onori, S. G. Corradini, P. Invernizzi, A. Franchitto, A. F. Attili, et al. Serum and Biliary Insulin-like Growth Factor I and Vascular Endothelial Growth Factor in Determining the Cause of Obstructive Cholestasis Ann Intern Med, October 2, 2007; 147(7): 451 - 459. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Razzaq, MBchB, MRCP, FRCR and S A Sukumar Imaging of the jaundiced adult Imaging, September 1, 2004; 16(4): 287 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Romagnuolo, M. Bardou, E. Rahme, L. Joseph, C. Reinhold, and A. N. Barkun Magnetic Resonance Cholangiopancreatography: A Meta-Analysis of Test Performance in Suspected Biliary Disease Ann Intern Med, October 7, 2003; 139(7): 547 - 557. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. R. Bader, K. L. Beavers, and R. C. Semelka MR Imaging Features of Primary Sclerosing Cholangitis: Patterns of Cirrhosis in Relationship to Clinical Severity of Disease Radiology, March 1, 2003; 226(3): 675 - 685. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Vitellas, A. El-Dieb, K. K. Vaswani, W. F. Bennett, M. Tzalonikou, C. Mabee, R. Kirkpatrick, and J. G. Bova MR Cholangiopancreatography in Patients with Primary Sclerosing Cholangitis: Interobserver Variability and Comparison with Endoscopic Retrograde Cholangiopancreatography Am. J. Roentgenol., August 1, 2002; 179(2): 399 - 407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Vitellas, A. El-Dieb, K. K. Vaswani, W. F. Bennett, J. Fromkes, C. Ellison, and J. G. Bova Using Contrast-Enhanced MR Cholangiography with IV Mangafodipir Trisodium (Teslascan) to Evaluate Bile Duct Leaks After Cholecystectomy: A Prospective Study of 11 Patients Am. J. Roentgenol., August 1, 2002; 179(2): 409 - 416. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Carlos, H. K. Hussain, J. H. Song, and I. R. Francis Gadolinium-Ethoxybenzyl-Diethylenetriamine Pentaacetic Acid as an Intrabiliary Contrast Agent: Preliminary Assessment Am. J. Roentgenol., July 1, 2002; 179(1): 87 - 92. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Vitellas, R. A. Enns, M. T. Keogan, K. S. Freed, C. E. Spritzer, J. Baillie, and R. C. Nelson Comparison of MR Cholangiopancreatographic Techniques with Contrast-Enhanced Cholangiography in the Evaluation of Sclerosing Cholangitis Am. J. Roentgenol., February 1, 2002; 178(2): 327 - 334. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Keogan and R. R. Edelman Technologic Advances in Abdominal MR Imaging Radiology, August 1, 2001; 220(2): 310 - 320. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
