| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University Hospital of Angers, 4, rue Larrey, Angers,
France 49933.
2 Department of Hepatogastroenterology, University Hospital of Amiens, Amiens,
France 80080.
3 Department of Hepatogastroenterology, University Hospital of Angers, Angers,
France 49933.
4 Department of Visceral Surgery, University Hospital of Angers, Angers, France
49933.
5 Department of Radiology, Amiens University, Amiens, France 80080.
Received January 19, 2004;
accepted after revision June 22, 2004.
Address correspondence to C. Aubé.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty-seven patients were prospectively enrolled. Inclusion criteria included acute pancreatitis, subclinical jaundice, and clinical features of common bile duct stone migration. Radial endoscopic sonography and MR cholangiopancreatography with the single-shot fast spin-echo technique were performed a maximum of 48 hr apart. The gold-standard diagnosis was obtained with ERCP (n = 20) or intraoperative cholangiography (n = 14) if the results of endoscopic sonography or MR cholangiopancreatography were abnormal or if a cholecystectomy was performed, or by clinical and biochemical follow-up (n = 11) if the results of endoscopic sonography and MR cholangiopancreatography were normal.
RESULTS. The final diagnosis was common bile duct stones in 16 patients, malignant obstructions in four, and another biliary disease in two (lithiasis migration aspect with papillary edema); 23 patients had no biliary disease. The sensitivity and specificity of MR cholangiopancreatography were, respectively, 90.5% and 87.5% for etiologic diagnosis and 87.5% and 96.6% for the detection of common bile duct stones. The corresponding values for endoscopic sonography were 86.4% and 91.3% for etiologic diagnosis and 93.8% and 96.6% for visualization of choledocholithiasis. Accuracy did not significantly differ between the techniques.
CONCLUSION. In cases of mild to moderate suspicion of choledocholithiasis, the accuracies of endoscopic sonography and MR cholangiopancreatography are similar. Because MR cholangiopancreatography is noninvasive, it may be preferred for this indication.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Many studies have confirmed the good performance of MR cholangiopancreatography in bile duct disease [4–10], particularly for the study of bile ducts within the liver and at the hilum [11, 12].
Although MR cholangiopancreatography does not have the same limitations as and is less invasive than endoscopic sonography, both techniques may be used to study the bile ducts. The overall view of the biliary tree provided by MR cholangiopancreatography seems to be better for investigating the convergence between the external biliary tree and intrahepatic bile ducts. However, the poor resolution in space still limits its performance for diagnosis of small calculi in the common bile duct.
The aim of this prospective study was to compare MR cholangiopancreatography and endoscopic sonography, using ERCP as a reference, in patients in whom lithiasis of the common bile duct was suspected and endoscopic sonography was indicated—that is, when the level of suspicion of obstruction by stones was mild or moderate. Strong suspicion of lithiasis was considered an immediate indication for ERCP.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patients
The study was performed between January 2000 and March 2002 on 47 patients
(25 women and 22 men) with a mean age of 59 years (range, 21–86 years).
Patients were enrolled in the gastroenterology department of two university
hospitals (Table 1). The major
inclusion criterion was a mild to moderate suspicion of bile duct obstruction
by lithiasis, indicating the need for endoscopic sonography. Patients in whom
tumoral obstruction was suspected (history of tumor, slowly progressive
biologic or clinical cholestasis, deteriorated performance status) or bile
duct obstruction was strongly suspected (visibility of common bile duct stones
on sonographic examination; common bile duct diameter >10 mm and associated
with gallbladder calculi on sonographic examination) were not included,
because in clinical practice these patients required ERCP for diagnosis and
treatment and no further morphologic examinations were necessary. Three groups
were defined according to clinical picture: In group 1 (n = 14)
biologic cholestasis or subclinical jaundice was the first presentation, with
no suspicion of tumoral disease, with or without accompanying cholangitis or
bile duct dilatation and with no obstruction identified by sonography; group 2
(n = 5) had acute pancreatitis with no history of chronic alcoholism;
and group 3 (n = 28) had abdominal pain associated with varying
degrees of cholestasis or cytolysis (defined, respectively, as an alanine
transaminase serum level >60 U/L or an aspartate transaminase serum level
>44 U/L) or a transitory increase in serum amylase, suggesting migration of
a stone.
|
Exclusion criteria were contraindications to MR cholangiopancreatography or endoscopic sonography. Two patients were excluded from the study because a therapeutic procedure was necessary between MR cholangiopancreatography and endoscopic sonography.
Among the 45 remaining patients, 22 had extrahepatic bile duct obstruction, originating from lithiasis in 16, tumor in four, and other causes in two (common bile duct stone migration with papillary edema). In group 1, 12 of the 13 patients had bile duct obstruction: 8 cases that were due to lithiasis and four to tumor. In group 2, three patients had bile duct obstruction: two cases due to lithiasis and one to edema of the papilla (which appeared to be migration of a stone and was classified as other cause). In group 3, only seven patients had bile duct obstruction: six cases due to lithiasis and one to edema of the papilla (appearing as migration of a stone and classified as other causes).
In four patients, obstruction was in the upper third of the common bile duct (18.2%); in two patients, in the middle third of the common bile duct (9.1%); in 15 patients, in the lower third of the common bile duct (68.2%); and in one patient, not specified (4.6%). The common bile duct was dilated in 13 patients, with a mean diameter of 11.1 mm (range, 8–15 mm; SD, 2.5 mm). Stones measured 6.8 mm on average (range, 1–13 mm; SD, 3.6 mm).
Methods
MR cholangiopancreatography was always performed first. Endoscopic
sonography was performed no more than 48 hr later.
MR cholangiopancreatography was performed with a 1.5-T MRI scanner (Signa Echo Speed; GE Healthcare) with a surface phased-array coil. Investigations were performed on fasting patients who were lying supine. Respiration was monitored to check the quality of periods of apnea. Adjustment using T2-weighted (TE = 90 msec) single-shot fast spin-echo axial slices was first followed by interlaced thick (20-mm) slices in the coronal and coronal oblique planes, using a single-shot fast spin-echo sequence with the following parameters: a TEeff of more than 800 msec; a TR of infinity; a matrix of 256 x 192, 256 x 256, or 320 x 256 according to the technologic upgrades of the system; an echo train of 104; and a field of view of 36–40 cm. This was followed by coronal oblique slices 5–7 mm thick with a TE of 110 msec, centered on the common bile duct. If necessary, strongly T2-weighted thin slices (TE >800 msec) and radial slices were obtained. When tumor obstruction was suspected, T1-weighted gradient-echo sequences were obtained after injection of gadolinium chelates. The investigation lasted between 20 and 30 min.
Endoscopic sonograms were obtained using an EUM 20 or EUM 30 apparatus (Olympus Optical Co.) with a radial technology ultrasound probe functioning at 7.5, 12, or 20 MHz. Tests were performed on fasting patients, under neuroleptanalgesia or general anesthesia, with the patient in a left lateral horizontal position.
The investigators who interpreted each type of investigation were unaware of the results of the other type. MR cholangiopancreatography was interpreted by a consensus of two investigators, both of whom were senior radiologists and one, a specialist in gastrointestinal diseases. There were two sets of investigators, one for each center. Endoscopic sonography was interpreted in real time by a unique gastroenterologist in each center who had not performed the ERCP.
The following data were obtained by imaging (endoscopic sonography and MR cholangiopancreatography): the normal or pathologic nature of the bile ducts (bile ducts were considered normal when the wall was normal and the diameter of the common bile duct did not exceed 7 mm or, in patients with a history of cholecystectomy, 10 mm [13]), the maximum diameter of the common bile duct, the site and cause (lithiasis, tumor, or other) of the obstruction, the number and size of calculi, and whether the investigation could be interpreted. The quality of the investigation was evaluated as good, average, or poor. The degree of certitude of the operator's diagnosis could be definite or could require confirmation.
If one of the two investigations (MR cholangiopancreatography or endoscopic sonography) had abnormal findings, ERCP (n = 20) or peroperative cholangiography (n = 3) was performed and became the diagnostic reference. When MR cholangiopancreatography and endoscopic sonography had normal findings, the gold standard was either peroperative cholangiography when a cholecystectomy was performed (n = 9), ERCP (n = 1), or clinical and laboratory follow-up for 3 months (n = 12). All patients whose gold standard was the last of these experienced no recurrence of pain or worsening of laboratory findings and were therefore considered to be true-negatives.
Statistical Analysis
Sensitivity, specificity, and positive and negative predictive values for
endoscopic sonography and MR cholangiopancreatography were calculated for the
following items: common bile duct dilatation, diagnosis of obstruction, cause
of obstruction (tumor, calculus, or other), and obstruction by lithiasis. MR
cholangiopancreatography and endoscopic sonography findings were compared by
calculating raw concordance (chi-square test) and real concordance (kappa
test), with continuity corrections when required. Links between two
quantitative variables were obtained by calculating the correlation
coefficient and the degree of significance (p < 0.05).
Means were expressed with their SD, and means and ranges were provided without any prejudgment of data distribution. Because of the few patients in the study, no subgroups were analyzed.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
MR Cholangiopancreatography
MR cholangiopancreatography results could be interpreted in 100% of
patients. Diagnoses were definite in 42 patients and required confirmation in
three. MR cholangiopancreatography showed obstructions to be due to lithiasis
in 15 patients (Figs. 1A,
1B,
2,
3A,
3B, and
4), to tumor in four patients,
and to other causes in three patients (stone migration with papillary edema,
duodenal diverticulum, and narrow end of common bile duct with no obvious
explanation).
|
|
|
|
|
|
In patients with obstruction shown by MR cholangiopancreatography to be due to lithiasis, the mean number of stones per patient was 2.1 (range, 1–5; SD, 1.57). The mean diameter of the largest stone was 7.4 mm (range, 1–13 mm; SD, 3.12 mm).
The sensitivity, specificity, and positive and negative predictive values of MR cholangiopancreatography for the diagnosis of calculi were, respectively, 87.5%, 96.55%, 93.33%, and 93.33%. For the diagnosis of tumor, all four indexes were 100%. The sensitivity, specificity, and positive and negative predictive values of MR cholangiopancreatography for the diagnosis of other causes of obstruction were, respectively, 50%, 95.35%, 33.33%, and 97.62%. The values for other diagnoses are listed in Table 2.
Endoscopic Sonography
Endoscopic sonography results could be interpreted in 100% of patients.
Diagnoses were definite in 43 patients and required confirmation in two.
Endoscopic sonography showed the obstruction to be due to lithiasis in 16
patients, to tumor in 3 patients, and to other causes in 2 patients (stone
migration and duodenal diverticulum).
In patients with obstruction shown by endoscopic sonography to be due to lithiasis, the mean number of stones per patient was 1.6 (range, 1–3 mm; SD, 0.82 mm). The mean diameter of the largest stone was 8.9 mm (range, 5–14 mm; SD, 3.16 mm).
The sensitivity, specificity, and positive and negative predictive values of endoscopic sonography for the diagnosis of calculi were, respectively, 93.75%, 96.55%, 93.75%, and 96.55%. The respective values were 75%, 100%, 100%, and 97.6% for the diagnosis of tumor and 50%, 97.67%, 50%, and 97.67% for the diagnosis of other causes of obstruction. The values for other diagnoses are listed in Table 2.
Disagreement Between Results for Lithiasis
In the lithiasis subgroup, one MR cholangiopancreatography result was
false-positive and two were false-negative. The first false-negative was for a
patient with multiple millimeter-sized fragments, located mainly in the middle
of the common bile duct (patient 4). The common bile duct was not dilated. In
this patient, endoscopic sonography revealed a stone but—by identifying
only one 7-mm stone impacted in the papilla, with slight dilatation of the
common bile duct— was discordant for size. The second false-negative was
for a patient with a 2-mm stone in the upper third of the common bile duct,
with no dilatation of the duct or intrahepatic bile ducts (patient 45). The
findings of endoscopic sonography and ERCP agreed in this patient for the
diagnosis of calculi but not for the calculi number. The false-positive was
for a patient with a 4-mm stone in the lower third of the common bile duct
with associated dilatation (patient 26). In this woman, endoscopic sonography
and ERCP performed 48 hr after MR cholangiopancreatography confirmed
dilatation of the common bile duct but did not show any stone
(Fig. 5). However, ERCP showed
papillary edema, suggesting stone migration.
|
In the lithiasis subgroup, one endoscopic sonography result was false-positive and one was false-negative. With ERCP, the false-negative was shown to be a 5-mm-diameter stone in the lower third of the common bile duct, with accompanying moderate (8-mm) dilatation (patient 22). In this patient, the findings of MR cholangiopancreatography and ERCP agreed in showing a stone smaller than 1 cm in a dilated common bile duct. The false-positive was a 14-mm stone falsely diagnosed in the upper third of a dilated common bile duct (patient 9). ERCP and MR cholangiopancreatography both showed a tumor in the upper third of the common bile duct, suggesting cholangiocarcinoma without marked bile duct dilatation. The presence of the tumor was confirmed by surgery. Other differences in diagnoses are shown in Table 3.
|
Results for Small Calculi (
5 mm)
Because the accuracies of the two techniques could be especially different
for the detection of small calculi, we identified patients with such calculi.
Of the six calculi 5 mm or less in diameter (range, 1–5 mm; SD, 1.68 mm)
discovered by ERCP, four were diagnosed by MR cholangiopancreatography with a
mean error in size assessment of 1 mm. The first misdiagnosis was in a patient
with a unique 2-mm calculus. The second was in a patient with numerous
millimeter-sized fragments. Five of the six calculi 5 mm or less in diameter
were diagnosed by endoscopic sonography with a mean error in size assessment
of 3 mm. The only misdiagnosis was in a patient with a unique 5-mm calculus.
Details of these false-negatives are reported in
Table 3.
Agreement Between Methods
Agreement was very good for diagnosing common bile duct dilatation, with
only three discordant results (
= 0.86), and good for diagnosing the
cause of obstruction, with five discordant results ( = 0.72). Agreement
between the two techniques for diagnosing intrahepatic bile duct dilatation
and identifying the site of obstruction was poor to moderate ( = 0.29
and = 0.63, respectively).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MR cholangiopancreatography, which requires no sedation, appears to be less expensive [16, 22] and much better tolerated, with patients' major complaint being a feeling of confinement while in the apparatus [23]. Limitations are the contraindications for MRI, although with the help of respiratory monitoring, 2 sec of apnea no longer seems to be essential to obtain good results [24]. This investigation can be standardized, allowing repeated interpretation. Performance in the diagnosis of intrahepatic and hilar biliary disease is good [11, 14, 24–27], whereas endoscopic sonography is limited by its lack of exploratory depth and ERCP can be limited when a complete obstruction prevents upstream opacification or when a decision is made to inject a smaller amount of contrast medium to avoid infection, such as in cases of sclerosing cholangitis [28].
For investigation of the common bile duct, the results of MR cholangiopancreatography can be compared with those of endoscopic sonography, which has no technical limitations in the investigation of this region. However, only 3 studies have compared these two techniques [16, 29, 30]. The literature reports an accuracy of between 90% and 97% for MR cholangiopancreatography in the diagnosis of common bile duct calculi [9, 31], but some authors have reported that results for stones less than 3 mm in diameter are poor with MR cholangiopancreatography [4, 16, 32].
The 3 studies that did compare MR cholangiopancreatography and endoscopic sonography found slightly better results with endoscopic sonography than did the present study, with sensitivities between 95% and 100% and specificities between 80% and 96%. MRI performance varies more. De Ledinghen et al. [30] reported good sensitivity (100%) but low specificity (62%) for MRI in the diagnosis of common bile duct lithiasis. The study of Materne et al. [29] showed 91% sensitivity and 94% specificity for MR cholangiopancreatography, close to the values for endoscopic sonography. The study of Scheiman et al. [16] showed significantly better results with endoscopic sonography (sensitivity, 95%; specificity, 80%) than with MR cholangiopancreatography (sensitivity, 40%; specificity, 96%) in the diagnosis of lithiasis obstruction of the common bile duct.
However, these studies investigated biliary disease in general, and the inclusion criterion was any suspected bile duct obstruction. Thus, the proportion of common bile duct stones was fairly low: 10 in the study of de Ledinghen et al. [30], eight in that of Materne et al. [29], and five in that of Scheiman et al. [16]. The diagnostic context was not the same as for our study, in which the indication was obstruction of the common bile duct by lithiasis, with a moderate probability of identifying obstructions. At present, this is the accepted indication for endoscopic sonography in disease of the common bile duct. The MRI technique in the study of Scheiman et al. used thin slices reconstructed later in maximum intensity projection, and these were probably less effective than thick slices [2, 32–34].
Our study had a number of limitations. First, patient inclusion was not consecutive because of the indication (mild or moderate suspicion of common bile duct stones) and the availability of a device within a 48-hr interval. Second, this 48-hr interval was in fact probably too long because of the acute and transient nature of some biliary diseases and the mobility of stones. Third, a question exists of whether ERCP is really the gold standard for the diagnosis of small stones in the common bile duct [11, 15], even though in this study endoscopists were aware that they should look for small stones, especially in the discharge of biliary fluid into the duodenum after sphincterotomy. At last, because ERCP was not performed when the findings of both endoscopic sonography and MR cholangiopancreatography were normal, small calculi that did not obstruct or dilate the common bile duct could easily have been missed. Moreover, in this case calculi may have spontaneously passed. However, although this situation would modify the accuracy of MR cholangiopancreatography and endoscopic sonography, it did not change the comparison between these techniques, which was the goal of this study.
Although the number of common bile duct stones investigated by endoscopic sonography and MR cholangiopancreatography in this study is, to our knowledge, the most reported in the literature, it still remains low. Nevertheless, the nearly equal distribution between patients with and without bile duct diseases made it possible to perform a robust statistical analysis of our results.
We found no significant difference in performance between endoscopic sonography and MR cholangiopancreatography. Of patients in whom a stone was present or suspected, MR cholangiopancreatography findings were false-negative for only two and false-positive for only one, and endoscopic sonography findings were false-negative for only one and false-positive for only one.
With MR cholangiopancreatography, the first false-negative concerned multiple 1-mm stones floating in the common bile duct. Endoscopic sonography, which was performed during the same anesthesia administration as for ERCP, detected a single 7-mm stone impacted in the papilla. This is a true-positive statistically but surprising morphologically.
The false-positive was a 4-mm stone that ERCP failed to detect. The MR cholangiopancreatography findings (Fig. 5) and the papillary edema found on ERCP suggested that the stone migrated and was evacuated between the time of MR cholangiopancreatography and the time of ERCP, 48 hr later.
One major limitation of MR cholangiopancreatography, as reported in the literature, is in the diagnosis of microcalculi [11, 16, 35]. In this study, no statistically significant difference was found between the two techniques for the diagnosis of small calculi. MR cholangiopancreatography detected four of the six calculi of 5-mm diameter or less, whereas endoscopic sonography detected five of the six if we include the surprising 7-mm stone that was found instead of the millimeter-sized fragments shown by ERCP. MR cholangiopancreatography appeared to be more accurate in determining the size of calculi.
The finding that was false-positive by both MR cholangiopancreatography and endoscopic sonography was a duodenal diverticulum. In view of the agreement between the two techniques and the moderate performance of ERCP for the diagnosis of duodenal diverticula [36], this case could be considered false-negative for ERCP.
In addition to the lack of difference in performance between the two techniques, agreement between them was good for all items noted except dilatation of the intrahepatic bile ducts. This moderate agreement could be explained by several factors: First, dilatation of the intrahepatic bile ducts was evaluated in a nonparametric manner (dilated or not), that is, leaving a degree of subjectivity to the operator. Second, endoscopic sonography uses high-frequency probes with very good resolution and only a shallow depth of exploration, and the intrahepatic bile ducts are therefore less well visualized. Third, the reference for this item could not be obtained because, to limit morbidity from ERCP [37], retrograde opacification was at low pressure and was stopped as soon as the obstruction was identified. Hence, the status of the intrahepatic bile ducts was not evaluated in most patients.
On the basis of the results of this study and the more general studies of de Ledinghen et al. [30] and Materne et al. [29], the minimum calculus size detectable by MR cholangiopancreatography cannot be defined. Although detection is limited by the use of a voxel of more than 1.4 mm along its shortest side, the powerful contrast between the hyperintensity of bile and the hypointensity of the stone compensates for this limitation by decreasing the calculus detectability thresh-old. However, this compensation requires the presence of bile around the calculus, that is, dilated bile ducts.
In cases of mild to moderate suspicion that the common bile duct is obstructed by lithiasis—the current indication for MR cholangiopancreatography and endoscopic sonography—the results for the two techniques do not significantly differ. MR cholangiopancreatography should be used in this situation because it is noninvasive, does not require sedation, is cheaper, provides good interobserver concordance [2], and is less operator-dependent. However, various parameters must be kept in mind: MRI equipment is still rare and thus not as available; MR cholangiopancreatography has certain difficulties of interpretation [38, 39]; and, depending on the skill of the main operator, endoscopic sonography may be preferred. It can also be done at the same time as ERCP, thus preventing the problem of sedation.
In conclusion, when a common bile duct calculus is mildly to moderately suspected (i.e., when a complementary morphologic examination is indicated before ERCP), MR cholangiopancreatography or endoscopic sonography can be the initial investigation. In this particular diagnostic situation, the present study found no difference between the results of endoscopic sonography and MR cholangiopancreatography. Thus, depending on equipment availability and operator skill, MR cholangiopancreatography may be preferable in these cases because it is noninvasive and less operator-dependent.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Stoker, A. van Randen, W. Lameris, and M. A. Boermeester Imaging Patients with Acute Abdominal Pain Radiology, October 1, 2009; 253(1): 31 - 46. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E J Williams, J Green, I Beckingham, R Parks, D Martin, and M Lombard Guidelines on the management of common bile duct stones (CBDS) Gut, July 1, 2008; 57(7): 1004 - 1021. [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]
|
|
|
|
 |
|
 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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
