| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Gifu University School of Medicine, 40 Tsukasamachi,
Gifu, Japan, 500-8705.
2
First Department of Internal Medicine, Gifu University School of Medicine,
Gifu, Japan, 500-8705.
Received December 15, 2000;
accepted after revision September 12, 2001.
Supported in part by the Grant-in-Aid for Cancer Research (12-4) from the
Ministry of Health, Labour and Welfare in Japan.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Unenhanced and enhanced MR cholangiopancreatograms obtained in 50 consecutive patients were quantitatively analyzed with region-of-interest measurements and were qualitatively evaluated by three independent radiologists unaware the patient information. A phantom study was performed to verify the effects.
RESULTS. The mean contrast-to-noise ratios of the gallbladder, common bile duct, and main pancreatic duct significantly increased after gadolinium chelate administration (p < 0.005). The mean depiction score of the main pancreatic duct increased significantly with one radiologist (p < 0.05) and marginally with another (p < 0.06), and the mean depiction scores of the background structures and renal pelvis significantly decreased with all three radiologists (p < 0.001). The phantom study showed the results, indicating that T2- and T2*-shortening effects of gadolinium chelate caused the effects in vivo.
CONCLUSION. IV administration of gadolinium chelate improves the depiction of pancreaticobiliary ducts in some selected patients, while decreasing the depiction in others with less frequency. There may be a value of enhanced MR cholangiopancreatography when unenhanced MR cholangiopancreatography is not sufficient.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Some researchers have described the usefulness of orally administered negative contrast agents to improve the quality of MR cholangiopancreatography [17, 18]. Meanwhile, Takahashi et al. [19] reported that gadolinium chelate administration suppressed the signals from splanchnic vessels with slow flow and improved the image quality of thick-section half-Fourier rapid acquisition with relaxation enhancement MR cholangiopancreatography. Takahashi et al. recommended performing MR cholangiopancreatography after administration of contrast material. However, we have experienced some anecdotal reports of patients in whom the correct diagnosis was improved after gadolinium chelate administration and of others in whom, contrarily, the diagnosis was obscured. The effects of IV administration of gadolinium chelate on the image quality of thick-section MR cholangiopancreatography, to our knowledge, are as yet uncertain. We quantitatively and qualitatively assessed the effects of gadolinium chelate on MR cholangiopancreatography performed with a thick-section half-Fourier rapid acquisition with relaxation enhancement sequence. We also performed phantom studies to verify the phenomenon.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MR imaging was performed with a 1.5-T MR unit (Signa Horizon, General Electric Medical Systems, Milwaukee, WI) with body phased array coil. The MR imaging protocol for this study consisted of unenhanced and enhanced thick-section MR cholangiopancreatography with fat-suppressed thick-section (20-50 mm) half-Fourier rapid acquisition with relaxation enhancement coronal and paracoronal imaging (effective TR/effective TE, infinite/1,034; echo-train length, 136; half-Fourier acquisition, 8 msec interecho spacing; matrix, 256 x 256; field of view, 24 x 24 cm; kHz received bandwidth, ±31.3; one location per 1 sec).
We performed unenhanced and enhanced thicksection MR cholangiopancreatography with the same pulse sequence and scan parameters. Enhanced imaging was begun immediately after completion of gadolinium-enhanced dynamic gradient-recalled echo axial imaging that was performed in all patients. For the gadolinium-enhanced gradient-recalled echo imaging, 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) was administered IV as a bolus. Enhanced MR cholangiopancreatography was performed between 5 and 10 min after initiation of gadolinium-chelate administration. Both unenhanced and enhanced thick-section MR cholangiopancreatography was performed in four locations (coronal, 30° and 60° right anterior oblique, and 30° left anterior oblique). The scan location and section thickness were identical with unenhanced and enhanced MR cholangiopancreatography.
Quantitative Image Analysis
Quantitative analysis was conducted on the unenhanced and enhanced
thick-section MR cholangiopancreatography using operator-defined
region-of-interest measurements of mean signal intensity of the visible
gallbladder, common bile duct, main pancreatic duct, and background structures
in three sites adjacent to the gallbladder, common bile duct, and main
pancreatic duct, respectively. Regions of interest for the background
structures were located devoid of gastrointestinal tracts, intrahepatic bile
ducts, focal hepatic lesions, or ascites, so that the site and area were the
same on unenhanced and enhanced MR cholangiopancreatography. Images were
arbitrarily magnified on the console monitor to place correctly regions of
interest. Mean signal intensity of the background structures in each patient
was obtained by averaging the signal intensities in the three sites
neighboring pancreaticobiliary ductal structures.
The contrast-to-noise ratio of each anatomic structure was calculated by dividing the difference in mean signal intensity between the anatomic structure and adjacent background structure by the standard deviation (SD) of the mean signal intensity of the adjacent background structure.
Qualitative Image Analysis
Three radiologists who worked as gastrointestinal radiologists and did not
know any information about the patients independently reviewed the unenhanced
and enhanced thick-section MR cholangiopancreatograms printed on films at the
same imaging windows and levels, in random order of patients and imaging
sequences. To minimize learning bias, the name, age, identification number,
and imaging parameters for each patient were masked.
Each radiologist subjectively scored images for depiction of the biliary tracts and main pancreatic duct in a 5-point scale, with a score of 1 assigned for no depiction; a score of 2, for poor depiction; a score of 3, for fair depiction; a score of 4, for good depiction; and a score of 5, for excellent depiction. A score of 5 was assigned when depiction was sufficient to recognize almost all ductal structures. A score of 3 was assigned when depiction was moderate, but recognition of the ductal structures was not precluded. A score of 1 was assigned when depiction was almost absent, and interpretation of the ductal structures was markedly precluded. Scores of 2 and 4 were assigned according to the radiologist's subjective judgment.
Each radiologist also subjectively scored images for depiction of the background structures and renal pelvis in a 3-point scale, with a score of 1 assigned for minimal or no depiction; a score of 2, for moderate depiction; and a score of 3, for intense depiction. A score of 3 was assigned when depiction was intense, and interpretation of the ductal structures was markedly precluded. A score of 2 was assigned when depiction was moderate, but recognition of the ductal structures was not precluded. A score of 1 was assigned when depiction was virtually absent.
Statistical Analysis
The means of signal intensities and contrast-to-noise ratios with
unenhanced and enhanced MR cholangiopancreatography were compared using the
paired Student's t test. The mean depiction scores with unenhanced
and enhanced MR cholangiopancreatography were compared using the Wilcoxon's
signed rank test.
To evaluate the degree of agreement with quantitative and qualitative analyses with unenhanced and enhanced MR cholangiopancreatography, correlation between changes in the contrast-to-noise ratio of the common bile duct and those in depiction scores for biliary tracts and correlation between changes in the contrast-to-noise ratio of the main pancreatic duct and those in depiction scores for the main pancreatic duct were tested using the Spearman's rank correlation coefficient.
To assess interobserver variability in assigning the depiction scores, multiple observer kappa statistics were used to measure the degree of agreement among the three radiologists. We used the nonweighted kappa statistic, with binary data defined in terms of the less-than-50% cutoff level. A kappa value of up to 0.20 stood for slight agreement; a value of 0.21-0.40, for fair agreement; a value of 0.41-0.60, for moderate agreement; a value of 0.61-0.80, for substantial agreement; and a value of 0.81 or greater, for almost perfect agreement.
Phantom Study
We prepared four agar phantoms to simulate thick-section MR
cholangiopancreatography before and after IV gadolinium chelate administration
in the human body. One was a plain agar phantom to simulate background
structures before contrast administration, and the other three were agar
phantoms with gadolinium chelate dissolved at concentrations of 0.1, 0.3, and
0.5 mmol/L to simulate background structures after contrast administration
[21]. A hematocrit tube
(1.4-mm caliber with 0.2-mm wall thickness) filled with normal saline solution
simulating the pancreaticobiliary duct was dipped in each agar phantom. The
phantoms were imaged with the same MR scanner, phased array multicoil, and
pulse sequence (effective TR/effective TE, infinite/1,034; section thickness,
50 mm) as those used for the clinical study.
The quantitative analysis was conducted using operator-defined region-of-interest measurements of mean signal intensity of the agar and hematocrit tubes. To simulate the contrast-to-noise ratio of pancreaticobiliary ducts in the human body, the contrast-to-noise ratio of the hematocrit tube was calculated by dividing the difference between the mean signal intensities of the agar and hematocrit tube by the SD of the mean signal intensity of the agar.
T2- and T2*-relaxation time of each agar phantom was calculated with the mean signal intensities measured on images obtained by using spin-echo (TR/TE, 2,000/40, 80, 120, 160) and gradient-recalled echo (500/10, 20, 30, 40; flip angle, 90°) sequences, respectively, with a head coil. The least square fitting was used to calculate the relaxation times.
To verify the effect of gadolinium chelate on the renal pelvis, one test tube filled with normal saline solution simulating the unenhanced renal pelvis and two filled with 10 and 20 mmol/L of gadolinium chelate solution simulating the enhanced renal pelvis were imaged with the same MR scanner, phased array multicoil, and pulse sequence, as those used for the clinical study.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
With thick-section MR cholangiopancreatography, the gallbladder was not visualized in 11 patients, normally depicted in 34, slightly to moderately distended in two, and markedly distended in three. The common bile duct, hepatic ducts, and intrahepatic bile ducts were normally depicted in 39 patients, slightly to moderately distended in seven, and markedly distended in four. The main pancreatic duct was not visualized or normally depicted in 44 patients, slightly to moderately distended in five, and markedly distended in one. The measurement was feasible for the gallbladder in 39 patients, the common bile duct in all, and the main pancreatic duct in 42.
The mean signal intensities for the gallbladder (not significant), common bile duct (p < 0.01), main pancreatic duct (not significant), and adjacent background structures (p < 0.001) decreased after gadolinium chelate administration (Table 1). Meanwhile, the mean contrast-to-noise ratios for the gallbladder (p < 0.005), common bile duct (p < 0.001), and main pancreatic duct (p < 0.001) increased after gadolinium chelate administration (Table 1).
|
The means of depiction scores for each anatomic structure subjectively judged by the three radiologists are summarized in Table 2. The mean depiction score for the biliary tracts did not significantly change for all three radiologists (Figs. 1A,1B,2A,2B,3A,3B,4A,4B,4C). The mean depiction score for the main pancreatic duct increased significantly for one radiologist (p < 0.05) and marginally for another (p < 0.06) (Figs. 1A,1B, 2A,2B, and 4A,4B,4C). The mean depiction score for the background structures and renal pelvis significantly decreased for all three radiologists (p < 0.001, identically) (Figs. 2A,2B and 3A,3B).
|
|
|
|
|
|
|
|
|
|
The transition of depiction scores assigned by the three radiologists from unenhanced to enhanced MR cholangiopancreatography is shown in Table 3. The depiction score of the biliary tracts increased after gadolinium chelate administration in four to 12 patients (range, 8-24%; mean, 15%) and decreased in four to 11 patients (range, 8-22%; mean, 13%), that of the main pancreatic duct increased in 10-18 patients (range, 20-36%; mean, 29%) and decreased in four to 10 patients (range 8-20%; mean, 13%), that of the background structures increased in zero to 1 patient (range, 0-2%; mean, 1%) and decreased in 26-37 patients (range, 52-74%; mean, 63%), and that of the renal pelvis increased in zero to 1 patient (range, 0-2%; mean, 1%) and decreased in 15-27 patients (range, 30-54%; mean, 41%).
|
There was a slight but significant correlation between changes in contrast-to-noise ratio of the common bile duct and those in depiction scores for biliary tracts with unenhanced and enhanced MR cholangiopancreatography in two radiologists (p = 0.33-0.37, p < 0.05). Likewise, there was a slight but significant correlation between changes in contrast-to-noise ratios for the main pancreatic duct and those in depiction scores for the main pancreatic duct in two radiologists (p = 0.34-0.38, p < 0.05).
The kappa values for the three radiologists were 0.48, 0.75, 0.58, and 0.61 for the evaluation of biliary tracts, main pancreatic duct, background structures, and renal pelvis, respectively. Substantial agreement among the observers was obtained for the main pancreatic duct and renal pelvis. Moderate agreement was obtained for the biliary tracts and background structures.
Phantom images are shown in Figure 5. The transitions of mean signal intensity of the agar and hematocrit tube, contrast-to-noise ratios of the hematocrit tube, and T2- and T2*-relaxation times of the agar for the concentration of gadolinium chelate in agar are shown in Figures 6,7,8, respectively. The test tubes filled with normal saline solution showed high signal intensity, but those filled with gadolinium chelate solution (10 and 20 mmol/L) almost completely disappeared. The quantitative results in the phantom study correlated well with the quantitative results in vivo.
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Gadolinium chelate is supposed to reach an equilibrium state in no more than 3-5 min after IV administration and is supposed to be distributed in vascular and extracellular spaces at an approximate concentration of 0.1-0.5 mmol/kg [21]. Our phantom study showed an evident signal intensity decrease of the agar presumably due to T2- and T2*-shortening effects of gadolinium chelate at these concentrations. Likewise, in our clinical study, the mean signal intensity of the background structures significantly reduced after gadolinium chelate administration.
The reasons for the reduced signal intensity of the background structures might be twofold: first, gadolinium chelate that reached an equilibrium state shortened the T2 relaxation time of extracellular fluid in the background structures. Second, gadolinium chelate disturbed the local magnetic field (T2*-shortening effect) and lowered the signal intensity of lymph or lymph vessels in the background structures. Because the rapid acquisition with relaxation enhancement sequence is supposed to be relatively resistant to the T2*-shortening effect because of its successive multiple 180° refocusing pulses with a short interecho space, the T2-shortening effect may be more influential than the T2*-shortening effect is.
IV administration of gadolinium chelate also decreased the signal intensity of the pancreaticobiliary ductal structures. However, the decrease was significant in the common bile duct alone. The reason for the significant signal intensity decrease exclusively in the common bile duct is unclear, but suspected cause is the susceptibility effect from the portal trunk that runs in the vicinity of the common bile duct. The portal blood flow containing gadolinium chelate at a presumed concentration of 0.5 mmol/L [21] might have disturbed the local magnetic field in the vicinity of the common bile duct and lowered the signal intensity of the common bile duct. In some patients, the main pancreatic duct or peripheral intrahepatic bile ducts became thinner after contrast administration (Fig. 3A,3B). This thinning may be similarly attributed to the susceptibility effect of gadolinium chelate that flowed in the neighboring pancreatic, splenic, or hepatic vessels or that was distributed in the ductal walls.
The renal pelvis containing urine was depicted as a structure of high signal intensity that might superimpose the pancreaticobiliary system on MR cholangiopancreatography, occasionally hampering the interpretations of MR cholangiopancreatography. A normal kidney starts excreting gadolinium chelate immediately after the first transit of contrast agent occurs, and the concentration of gadolinium chelate in urine is fairly high (10-40 mmol/L) at that time, resulting in extremely short T1 and T2 values (<30 msec) [22] and in the disappearance of the renal pelvis on MR cholangiopancreatography. This clinical observation was also verified by the phantom study.
The clinical benefit of increased delineation of the main pancreatic duct produced by IV gadolinium-chelate administration might be that the diagnosis of anomalous pancreaticobiliary ductal union is improved as was shown in one patient in our study (Fig. 4A,4B,4C). Recent studies have reported that a variety of pancreaticobiliary disease in children and adolescents is associated with this congenital disorder [24,25,26,27,28,29], and gastroenterologists warn that anomalous pancreaticobiliary ductal union should always be remembered in treating patients with abdominal pain and gallbladder wall thickening that do not accompany gallstones, dilatation of the common bile duct, or biliary tumor [26]. MR cholangiopancreatography has limited value in delineating undistended pancreaticobiliary ducts because of the limited spatial resolution, artifacts, or patient motion, and endoscopic retrograde cholangiopancreatography is often mandated when anomalous pancreaticobiliary ductal union is suspected [25, 26]. We assume that our current results may indicate a potential usefulness in diagnosing this congenital disorder. Although secretin-stimulated MR cholangiography may be promising in the diagnosis of anomalous pancreaticobiliary ductal union or pancreatic divisum [30], its practical usefulness has been as yet uncertain.
We observed an unusually increased visualization of the main pancreatic duct after gadolinium chelate administration in one patient (Fig. 1A,1B), in whom duodenojejunal fluid that had been minimally seen on unenhanced MR cholangiopancreatography was obviously increased on enhanced MR cholangiopancreatography. The reason was unclear, but gadolinium chelate is known to affect the autonomic nervous system, possibly leading to adverse gastrointestinal reactions such as nausea or vomiting [31]. There may be, as yet, unknown effects of gadolinium chelate on stimulation of the pancreatic exocrine function.
There are some limitations to this study. To a certain extent, it elucidated the effects of gadolinium chelate on image quality of MR cholangiopancreatography, but the real impact of these effects on the diagnosis of pancreaticobiliary disease is as yet undetermined. Observer performance studies on how enhanced MR cholangiopancreatography improves the diagnosis are needed. We did not use an oral negative contrast agent for elimination of signals of the gastrointestinal tracts [17, 18] because we tried to visualize the duodenum for better diagnosing duodenal papillary tumor, pancreatic divisum, or anomalous pancreaticobiliary ductal union. We need to assess further the interaction of IV administered gadolinium chelate and orally administered negative contrast material.
In conclusion, an IV administration of gadolinium chelate frequently suppresses the depiction of the background structures and renal pelvis and improves the depiction of the biliary tracts or main pancreatic duct in selected patients. Meanwhile, with less frequency, the depiction of the pancreaticobiliary ducts can deteriorate after administration of gadolinium chelate. Gadolinium chelate administration for the purpose of improving diagnosis with MR cholangiopancreatography may be as yet unwarranted, and prospective studies in larger populations are necessary. At present, however, there may be a value of enhanced thick-section MR cholangiopancreatography when the diagnostic value of unenhanced MR cholangiopancreatography is not sufficient.
Acknowledgments
We thank Tetsuji Kurata of General Electric Yokogawa Medical Systems and
Satoshi Yoshise of Nihon Schering for technical advice.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
