| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Hepatobiliary Imaging |
1 Department of Diagnostic Radiology, Chonbuk National University Hospital,
Conju, South Korea.
2 Department of Radiology and Institute of Radiation Medicine, Seoul National
University Hospital, 28, Yongondong, Chongno-gu, Seoul 110-744, South
Korea.
Received January 5, 2004;
accepted after revision March 19, 2004.
Address correspondence to J. M. Lee
(leejm{at}radcom.snu.ac.kr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Forty-six patients with 96 hepatocellular carcinomas underwent Mn-DPDP– and gadopentetate dimeglumine–enhanced MRI. The MRI examination included unenhanced T2-weighted turbo spin-echo and T1-weighted 2D fast low-angle shot (FLASH) sequences and a 3D FLASH sequence after the administration of gadopentetate dimeglumine and Mn-DPDP. Two observers reviewed three sets of images: a set of gadopentetate dimeglumine–enhanced MR images, a set of Mn-DPDP–enhanced MR images, and a combination of the gadopentetate dimeglumine and Mn-DPDP sets. Using receiver operating characteristic (ROC) analysis, imaging sets were compared for diagnostic accuracy and sensitivity.
RESULTS. The area under the ROC curve (Az) was 0.942 for the gadopentetate dimeglumine–Mn-DPDP set, 0.932 for the gadopentetate dimeglumine set, and 0.877 for the Mn-DPDP set (p < 0.05). The mean sensitivity was greater for the gadopentetate dimeglumine set than for the Mn-DPDP set (87.5% vs 72.4%; p < 0.05). The false-negative rate of the Mn-DPDP set was statistically greater than that of the gadopentetate dimeglumine set (27.6% vs 12.5%). Most false-negative cases in the Mn-DPDP set were due to small (diameter < 2 cm), isoenhanced lesions.
CONCLUSION. Gadopentetate dimeglumine–enhanced MRI was superior to Mn-DPDP–enhanced MRI for the detection of hepatocellular carcinomas.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
With the recent development of liver-specific contrast agents such as superparamagnetic iron oxides and mangafodipir trisodium (Mn-DPDP), the uptake of contrast agents by functional cells has also increased the sensitivity of lesion detection and the specificity for tissue characterization [11–16]. Mn-DPDP is a paramagnetic hepatobiliary contrast agent taken up by the functioning hepatocytes and eliminated mainly by the biliary tract; it has shown its efficacy for improving the detection and delineation of liver lesions in comparison with unenhanced imaging [17–19]. To our knowledge, only one study [20] has compared gadopentetate dimeglumine–enhanced MRI and Mn-DPDP–enhanced MRI for the detection of focal liver malignancies. The comparison of these two agents for the detection of focal liver malignancies, using currently available MRI units, seems to be of clinical and financial importance.
The purpose of this study was to compare the performance of Mn-DPDP–enhanced and gadopentetate dimeglumine–enhanced MRI for the detection of hepatocellular carcinoma.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Lesion Confirmation
Confirmation of the diagnosis and determination of the number and size of
the hepatic lesions as a standard of reference in receiver operating
characteristic (ROC) analysis were based on-site by means of interpretation of
all available data in each patient by two experienced radiologists in
consensus; these radiologists, who served as study coordinators, did not take
part in the ROC analysis. These data included histologic analysis, all imaging
data (sonography, dynamic helical CT, MRI, iodized oil–enhanced CT
obtained 2 weeks after transcatheter arterial chemoembolization, digital
subtraction angiography with CT arteriography, and CT portography), clinical
data, laboratory data, and follow-up findings for a minimum of 6 months and a
maximum of 1.5 years.
The diagnosis of hepatocellular carcinoma was based on histologic
examination in 17 patients; in the remaining 29 patients, the diagnosis was
based on clinical laboratory data (positive findings of hepatitis B surface
antigen or hepatitis C antibody and serum 
1-fetoprotein
level of > 200 ng/mL) and CT findings during hepatic arteriography and
arterial portography, findings at MRI, or both or in combination with typical
angiographic findings during transcatheter arterial chemoembolization
(n = 9), or specific findings on iodized-oil–enhanced CT
(n = 20).
MRI Examinations
All MRI examinations were performed on one of two 1.5-T MRI systems
(Magnetom Vision or Magnetom Symphony, Siemens) using a phased-array
multicoil. Unenhanced images included T2-weighted respiratory-triggered turbo
spin-echo sequences (TR range/TE, 2,800–3,200/101; acquisition time, 16
sec; flip angle, 150°; matrix, 136 x 256) and T1-weighted fast
low-angle shot (FLASH) sequences (TR/TE, 128/4.1 or 6.0; acquisition time, 17
sec; flip angle, 70° or 80°; matrix, 136 x 256) with 7-mm
section thickness and a 10% gap. Multiphasic dynamic T1-weighted 3D FLASH
sequences with fat saturation (3.4/1.6; acquisition time, 20 sec; flip angle,
20°; matrix, 121 x 256) were acquired before and 20, 40, and 120 sec
after the rapid bolus injection of gadopentetate dimeglumine (Magnevist,
Schering) at a dose of 0.1 mmol/kg using an automated injector (Medrad). The
injection was followed by a 20-mL saline flush.
Mn-DPDP–enhanced MRI was performed after the infusion of 0.5 mL/kg of Mn-DPDP (Teslascan, Amersham Health) during 10 min. Fifteen to thirty minutes after the administration of the contrast agent, T1-weighted MR images were obtained using 3D FLASH and fat saturation (3.8/1.8; acquisition time, 20 sec; flip angle, 15°; matrix, 115 x 256) and 2D FLASH with and without fat saturation (166/3.8; acquisition time, 19 sec; flip angle, 70° or 80°; matrix, 115 x 256) with 7-mm section thickness and a 10% gap. The field of view was 240–270 x 300–360 mm. All patients underwent Mn-DPDP–enhanced MRI after gadopentetate dimeglumine–enhanced MRI.
MR Image Analysis
The retrospective reviewing procedure in the ROC analysis was performed
independently in three sessions by two observers with experience in abdominal
MRI without any clinical information or diagnosis. To limit the learning bias,
the interval between sessions was at least 2 weeks, and images of one group
were randomly assigned to each observer and to each session. In the first
session, the two observers reviewed a set of images (gadopentetate dimeglumine
set) that included both unenhanced and gadopentetate
dimeglumine–enhanced images. In the second session, each observer
reviewed a set of images (Mn-DPDP set) that included both unenhanced and
Mn-DPDP–enhanced images of each lesion. In the third session, the
observers reviewed images of the Mn-DPDP sets and the gadopentetate
dimeglumine sets together. The images of all unenhanced and contrast-enhanced
sequences (gadopentetate dimeglumine–Mn-DPDP set) were presented to each
observer simultaneously, and the combined assessment of all images was used
for scoring. Each observer recorded the number of suspected lesions seen,
their size, and the segmental location of the lesions according to the
classification scheme of Couinaud, and assigned each lesion a confidence
rating score for the presence of hepatocellular carcinoma based on a 4-point
scale as follows: 5, definitely present; 4, probably present; 3, undetermined;
2, probably not present; and 1, definitely not present.
In addition, two radiologists who did not take part in the ROC analysis evaluated all lesions that were not identified or pseudolesions that were identified by any observer on the images of the gadopentetate dimeglumine set or the Mn-DPDP set for potential explanations about why the lesions were missed or the pseudolesions were detected.
Statistical Analysis
For all image sets, alternative-free-response ROC curves were plotted for
all lesions. ROC evaluation software (ROCKIT 0.9B, Charles E. Metz) was used,
and the diagnostic accuracy of each image set was determined by calculating
the area under the ROC curve (Az). Differences among ROC
curves were tested for significance (p < 0.05) using the
two-tailed area test for paired data. The lesions assigned a score of 4 or 5
by each reviewer were regarded as correctly diagnosed lesions, and sensitivity
values were calculated on this basis. Also, false-positive and false-negative
values were obtained for each image set. Differences in calculated values in
the detection of hepatocellular carcinoma were tested for significance
(p < 0.05) by performing the two-tailed Student's t
test.
The kappa statistic was used to measure the degree of agreement among the observers, and kappa values greater than 0 were considered to indicate positive correlation. Measurements were rated as follows: kappa values of 0.31–0.60, good correlation; 0.61–0.90, very good correlation; and greater than 0.90, excellent correlation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
Sensitivity and False-Positive and False-Negative Values
Table 2 summarizes the
sensitivity and false-positive and false-negative values of the
contrast-enhanced image sets in the detection of hepatocellular carcinoma
determined using the results of ROC analysis on the basis of scores of 4 or 5
assigned by the two observers. The best sensitivity was achieved for the
gadopentetate dimeglumine–Mn-DPDP set (88.0%). The sensitivity of both
the gadopentetate dimeglumine–Mn-DPDP set and the gadopentetate
dimeglumine set (87.5%) was significantly superior to that of the Mn-DPDP set
(72.4%) (p < 0.05).
|
The false-positive rate of the gadopentetate dimeglumine set was statistically greater than that of the Mn-DPDP set. The main cause of false-positive cases in the gadopentetate dimeglumine set was an arterioportal shunt (69%, 31/45) (Figs. 2A, 2B, and 2C). The false-negative rate of the Mn-DPDP set was statistically greater than that of the gadopentetate dimeglumine set. Most false-negative cases in the Mn-DPDP set were the result of a small (diameter < 2 cm), isoenhanced lesion (83%, 19/23) (Figs. 3A, 3B, and 3C).
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mn-DPDP is a hepatocyte-specific contrast agent capable of increasing the signal intensity of normal liver. At present, the best clinical role for this agent appears to be in improving detection of the number and extent of focal liver lesions in patients in whom hepatic resection is being contemplated [15, 26]. Also, Mn-DPDP–enhanced MRI can characterize lesions to a greater degree than unenhanced imaging can—that is, tumors of hepatocytic origin, such as hepatocellular carcinoma, hepatic adenoma, or focal nodular hyperplasia, that retain sufficient hepatocyte function actually take up the contrast agent and appear brighter on T1-weighted images than those of nonhepatocytic origin, such as metastases, cholangiocarcinomas, and cysts [15, 18]. But gadopentetate dimeglumine–enhanced MRI is used for routine MRI examination of the liver, and the most compelling case for the routine use of gadopentetate dimeglumine chelates is examination of the liver in patients who are thought to have hepatocellular carcinoma [27].
In our study, using ROC analysis we attempted to compare the diagnostic accuracy of Mn-DPDP–enhanced MRI with that of dynamic gadopentetate dimeglumine–enhanced MRI for the detection of hepatocellular carcinoma. For the diagnosis of hepatocellular carcinoma, the gadopentetate dimeglumine set was significantly more accurate and more sensitive than the Mn-DPDP set (p < 0.05). These results agreed with the results of the previous report by Kettritz et al. [20].
Several factors make gadopentetate dimeglumine–enhanced MRI superior to Mn-DPDP–enhanced MRI for diagnostic accuracy in hepatocellular carcinomas. First, depending on the amount of uptake of the contrast agent, hepatocellular carcinoma may become hyperintense, hypointense, or isointense on Mn-DPDP–enhanced MR images [16]. If the image shows greater or less enhancement than liver parenchyma, or an obvious hypointense capsule on Mn-DPDP–enhanced MRI, hepatocellular carcinoma can be reliably identified [18]. However, because of their enhancement that is as much as that of background liver parenchyma and that becomes isointense, some hepatocellular carcinomas, particularly small lesions, may be obscured on Mn-DPDP–enhanced MRI, which would make the false-negative rate for Mn-DPDP–enhanced MRI greater than that for gadopentetate dimeglumine–enhanced MRI [16] (Figs. 3A, 3B, 3C). Second, although the ultimate usefulness of Mn-DPDP for detecting hepatocellular carcinoma in patients with cirrhosis remains to be determined, some authors have reported that benign hepatocellular lesions, including regenerating nodules without iron deposits, may have homogeneous, hyperintense enhancement simulating that of some hepatocellular carcinomas, and it can be difficult to differentiate these benign lesions from hepatocellular carcinoma after the administration of Mn-DPDP [10, 13–15, 17, 28, 29].
However, Mn-DPDP–enhanced MRI was helpful in decreasing the false-positive diagnoses (Table 2). Most of the false-positives on gadopentetate dimeglumine–enhanced MRI seemed to result from perfusion abnormalities such as an arterioportal shunt of the liver parenchyma. On Mn-DPDP–enhanced MRI, the arterioportal shunt was not depicted and the liver parenchyma showed relatively homogeneous enhancement (Figs. 2A, 2B, and 2C). Also, Mn-DPDP successfully shows improved conspicuity and delineation of many liver lesions [16–18].
We also assessed the performance of a combined gadopentetate dimeglumine–Mn-DPDP set for its diagnostic accuracy using ROC analysis. The diagnostic accuracy of the gadopentetate dimeglumine–Mn-DPDP set was the best, and the false-positive value of the gadopentetate dimeglumine–Mn-DPDP set was as low as that of the Mn-DPDP set. From the results, Mn-DPDP–enhanced MRI is expected to have an axillary role for the diagnosis of hepatocellular carcinoma.
Our study has a number of limitations. First, the lesions of 29 patients were not pathologically proven. To have the best achievable gold standard, typical clinical and laboratory findings, in combination with typical findings of other imaging techniques, are used as the diagnostic criteria, but careful follow-up was performed in these patients. However, the possibility cannot be totally excluded that not all the lesions actually existed or, if they existed, were hepatocellular carcinomas. Also, we do not know about tumoral differentiation in hepatocellular carcinomas, which would be helpful for analyzing our results because the degree of enhancement on Mn-DPDP–enhanced MRI might be related to the differentiation of hepatocellular carcinomas. Second, many patients had hepatic dysfunction that might limit the degree of enhancement of the liver parenchyma and lower the diagnostic efficacy of Mn-DPDP. Last, because the study was performed retrospectively and excluded benign lesions such as hemangiomas or cysts from the analysis, a selection bias may exist and may limit the extrapolation of the data of this study to clinical situations.
In conclusion, multiphase dynamic gadopentetate dimeglumine–enhanced MRI was superior to Mn-DPDP–enhanced MRI; Mn-DPDP–enhanced MRI had only a limited role in the diagnosis of hepatocellular carcinoma.
Acknowledgments
We thank Bonnie Hami, Department of Radiology, University Hospitals of
Cleveland, Cleveland, OH, for her editorial assistance and manuscript
preparation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T. C. Lauenstein, K. Salman, R. Morreira, T. Heffron, J. R. Spivey, E. Martinez, P. Sharma, and D. R. Martin Gadolinium-Enhanced MRI for Tumor Surveillance Before Liver Transplantation: Center-Based Experience Am. J. Roentgenol., September 1, 2007; 189(3): 663 - 670. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
),
gadopentetate dimeglumine set (
), and Mn-DPDP set (
).
