| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiology and Institute of Radiation Medicine, Seoul National University Hospital, 28, Yongon-dong, Chongno-gu, Seoul 110-744, Korea.
Received October 25, 2006;
accepted after revision June 19, 2007.
Address correspondence to J. M. Lee
(leejm{at}radcom.snu.ac.kr;
jmlshy{at}naver.com).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The CT findings on 28 lesions in 26 patients with underlying liver cirrhosis and pathologically proven cholangiocarcinoma were retrospectively evaluated. The CT findings of hepatocellular carcinoma (HCC) in a control group of 79 subjects also were analyzed. The relative attenuation and enhancement pattern of the lesions were evaluated by two observers in consensus. The difference between the enhancement pattern of cholangiocarcinoma and that of HCC was statistically analyzed with the Fisher's exact test.
RESULTS. The prevalent enhancement patterns of cholangiocarcinoma on enhanced CT scans differed depending on tumor size. Peripheral rimlike enhancement was the most frequent (nine of 20 cases) pattern in tumors larger than 3 cm in diameter. A washout pattern on portal venous phase scans was the most frequent (five of eight cases) in tumors smaller than 3 cm in diameter. For tumors smaller than 3 cm in diameter, there was no significant difference between the enhancement pattern of cholangiocarcinoma and that of HCC. For tumors larger than 3 cm, the presence of peripheral rimlike enhancement or centripetal enhancement and the absence of a washout pattern were significant findings for differentiating cholangiocarcinoma from HCC (p < 0.0001).
CONCLUSION. The contrast enhancement patterns of cholangiocarcinoma in cirrhotic liver on multiphasic helical CT scans were found to differ depending on tumor size. Because of the overlapping imaging findings in the two diseases, for any hypovascular lesion smaller than 3 cm in a cirrhotic liver, the diagnosis of cholangiocarcinoma should be seriously considered along with that of HCC.
Keywords: cholangiocarcinoma • CT • hepatocellular carcinoma • liver cirrhosis • liver neoplasms
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Contrast-enhanced dynamic CT has a primary role in the differential diagnosis of focal liver lesions, including HCC, cholangiocarcinoma, metastatic lesions, and hemangioma [10–12]. Although it is possible to arrive at a differential diagnosis of a focal hepatic lesion with multiphasic helical CT in many cases of liver tumors, there is overlap between HCC and cholangiocarcinoma. However, given that there is a high incidence of HCC in cirrhotic liver and that HCC can have a variety of CT enhancement patterns [13–16], any focal solid hepatic lesion in cirrhotic liver can be easily regarded as HCC by many radiologists and physicians. There is a risk that this misdiagnosis can then lead to inappropriate use of treatments, such as transarterial chemoembolization, aimed at HCC without pathologic confirmation.
Although some reports [6, 8, 17] have described the clinicopathologic characteristics of intrahepatic cholangiocarcinoma in cirrhotic liver, to our knowledge there has never been a report describing the imaging findings of intrahepatic cholangiocarcinoma in cirrhotic liver. The purposes of this study were to assess the CT features and enhancement patterns of intrahepatic cholangiocarcinoma in cirrhotic liver and to evaluate the enhancement patterns of dual-phase helical CT for differentiating cholangiocarcinoma from HCC in cirrhotic liver.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An additional group of 79 consecutively enrolled control subjects (65 men, 14 women; mean age, 56.2 years; range, 26–83 years) had HCC identified at surgical resection between January 2006 and December 2006. These patients were selected according to the following criteria: they had diagnosed HCC and liver cirrhosis according to the surgical specimen; they had no previous diagnosis of or therapy for HCC; they had available dual-phase helical CT scans; the type of HCC was nodular or mass forming; and the patients had Child-Pugh class A or B disease (most of the patients with Child-Pugh class C disease could not undergo tumor resection). In the control group, the underlying causes of liver cirrhosis were hepatitis B (n = 66), alcohol abuse (n = 1), hepatitis C (n = 8), and unknown cause (n = 4). Cirrhosis was clinically classified as Child-Pugh class A in 74 patients and class B in five patients. Histopathologic examination of the lesion specimens revealed a total of 79 HCCs in these 79 patients.
CT
CT was performed on all 105 patients in the study and control groups and
included acquisition of contrast-enhanced images of all 105 patients and
unenhanced images of 95 patients (cholangiocarcinoma group, n = 21;
HCC group, n = 74). CT was performed at our institution on a
single-detector unit (Somatom Plus-4, Siemens Medical Solutions) or on an MDCT
unit (MX8000, Philips Medical Systems; Lightspeed Ultra, GE Healthcare;
Sensation 16, Siemens Medical Solutions; Brilliance 64, Philips Medical
Systems). Among the 105 study patients, 27 underwent CT on a single-detector
unit; the others underwent 4-MDCT (n = 14), 8-MDCT (n = 18),
16-MDCT (n = 37), and 64-MDCT (n = 9).
Single-detector helical CT was performed with the following parameters: 5-mm collimation, 5-mm reconstruction interval, and 1:1 table pitch. The parameters for MDCT were as follows: detector collimation, 0.75–2.5 mm; table speed, 20–24 mm/s; effective section thickness, 2.5–3.2 mm; reconstruction interval, 2.5–3.0 mm; gantry rotation time, 0.5–0.7 second; 170–220 effective mAs; 120 kVp; and 512 x 512 matrix.
All patients underwent dual-phase helical CT that included hepatic arterial phase (HAP) and portal venous phase (PVP) imaging begun 35 and 65 seconds, respectively, after IV infusion of 120 mL of nonionic contrast material (iopromide, Ultravist 370, Bayer HealthCare). A mechanical power injector (LF-9000, Liebel-Flarsheim) was used to administer the contrast material at a rate of 3 mL/s through an 18-gauge angiographic catheter inserted into a forearm vein. Transverse images were reconstructed with a soft-tissue algorithm.
Image Analysis
The CT images were reviewed retrospectively and jointly by the interpreting
radiologists (12 and 7 years of experience) in a blinded manner. These
reviewers knew that the patients had liver tumors and liver cirrhosis but did
not know the histopathologic diagnosis or results of the laboratory tests. All
CT scans were reviewed in the stack mode on a PACS workstation (Marosis,
Marotech). The observers attempted to determine the number, size, margin, and
enhancement patterns of the tumors. The lesion margin was defined according to
the following classifications: sharp and round, lobulated, and ill-defined.
The reviewers also assessed the presence of the following five morphologic
features on CT: capsular retraction, pseudocapsule, bile duct dilatation,
arterioportal shunt or transient hepatic attenuation difference (THAD), and
portal vein involvement including portal vein thrombosis and portal vein
invasion. Differences between the observers were resolved by consensus
conference.
The attenuation of the lesions compared with that of the surrounding liver parenchyma was classified as isoattenuating, hypoattenuating, or hyperattenuating at soft-tissue window settings (width, 400 H; level, 70–80 H) during the unenhanced and enhanced phases. Lesions that had an inhomogeneous enhancement pattern on the contrast-enhanced images were categorized as isoattenuating, hyperattenuating, or hypoattenuating according to the attenuation of the predominant parts of lesions. To ensure accurate classification of the relative lesion attenuation, CT numbers were obtained with region-of-interest cursors placed on the lesions and on the liver parenchyma. A difference of more than 10 H between the tumor and the liver attenuation was considered significant [18].
The lesion enhancement patterns were grouped on the HAP scans according to the following classification: peripheral rim enhancement, peripheral and central enhancement, and minimal enhancement. With reference to the enhancement features of the lesions on the HAP images, the lesion enhancement patterns on the PVP scans were grouped according to the following classification: peripheral rim enhancement, centripetal pattern, minimal enhancement, washout pattern, and persistent hyperattenuating enhancement.
Statistical Analysis
The difference between the enhancement pattern of cholangiocarcinoma and
that of HCC was statistically analyzed with the Fisher's exact test with
InStat software (GraphPad Software). A statistically significant difference
was p < 0.05. The numbers used in the statistical analysis were
numbers of lesions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Relative Tumor Attenuation
The results of subjective analysis showed that all 23 cholangiocarcinomas
on the 21 unenhanced scans were either hypoattenuating (20 lesions, 87%) or
isoattenuating (three lesions, 13%). On the HAP scans, six (21.4%) of the 28
lesions became hyperattenuating, and three (10.7%) became isoattenuating. The
other 19 (67.9%) lesions remained predominantly hypoattenuating. On the PVP
scans, all lesions became either hypoattenuating (27 lesions, 96.4%) or
isoattenuating (one lesion, 3.6%). Among the 28 cholangiocarcinomas,
enhancement of 11 (39.3%) of the lesions was relatively homogeneous, and that
of 17 (60.7%) of the lesions was inhomogeneous
(Fig. 1).
|
In the HCC group, most (61 lesions, 82.4%) of the 74 HCCs visualized on unenhanced scans were either hypoattenuating or isoattenuating (12 lesions, 16.2%). Only one lesion (1.4%) appeared hyperattenuating owing to hemorrhagic change. On the HAP scans of the 79 lesions, 65 (82.3%) of the lesions became hyperattenuating, and nine (11.4%) of the lesions became isoattenuating. The other five lesions (6.3%) remained predominantly hypoattenuating. On the PVP scans, 60 (75.9%) of the lesions became hypoattenuating, and 12 (15.2%) of the lesions became isoattenuating. The other seven (8.9%) lesions remained predominantly hyperattenuating. Among the 79 HCCs, 21 (26.6%) of the lesions had relatively homogenous enhancement, and 58 (73.4%) had inhomogeneous enhancement (Fig. 1).
Enhancement Patterns
In the cholangiocarcinoma group, the prevalent tumor enhancement patterns
on the HAP and PVP scans differed according to tumor size. Twelve (60%) of the
20 tumors larger than 3 cm in diameter had peripheral rimlike enhancement on
HAP or PVP scans, and nine (45%) had centripetal enhancement on the PVP scans
(Figs. 2A,
2B, and
2C). However, four (50%) of the
eight tumors smaller than 3 cm in diameter had minimal contrast enhancement on
the HAP scans, and five (62.5%) had a washout pattern on PVP scans (Figs.
3A,
3B,
4A, and
4B). In the HCC group, however,
there was no significant difference in the prevalent lesion enhancement
patterns according to tumor size; that is, a washout pattern (Figs.
5A and
5B) was visualized in 19
(86.4%) of the 22 small lesions and in 44 (77.2%) of the 57 large lesions.
|
|
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In clinical practice, CT and MRI are the only useful diagnostic methods in cases in which levels of tumor markers are not elevated and the biopsy results are indeterminate. Furthermore, for nodules larger than 2 cm in diameter in patients with cirrhosis, radiologists have a tendency to prematurely establish the diagnosis of HCC on imaging studies before obtaining pathologic confirmation [21, 22]. Because misdiagnosis of cholangiocarcinoma as HCC can lead to inappropriate therapy, our results have important practical value. In HCC patients, treatment is complete resection if the liver is minimally cirrhotic and liver transplantation if cirrhosis is more advanced. However, two thirds of patients with HCC are not good candidates for surgical resection because they have advanced cirrhosis or advanced HCC. They therefore undergo palliative treatments such as transarterial chemoembolization and systemic chemotherapy [21, 23]. In contrast, the optimal treatment of patients with cholangiocarcinoma is complete tumor resection including complete lymph node removal [4, 24, 25]. In cholangiocarcinoma patients with cirrhosis of the liver, complete tumor resection can be difficult in the presence of advanced cirrhosis, and liver transplantation is controversial because of the poor prognosis of cholangiocarcinoma. On the basis of our observations, we suggest that imaging-guided biopsy be performed in cases of small nodules in a cirrhotic liver.
Several other imaging features can be useful for identifying cholangiocarcinoma in cirrhotic liver. Capsular retraction, bile duct dilatation, THAD, and portal vein invasion were more frequently found in cholangiocarcinomas within cirrhotic liver than in HCC. In previous studies [26–29], THAD and portal vein invasion have been frequently associated with HCC. However, cholangiocarcinoma has a periportal growth pattern, and portal vein invasion and THAD are relatively frequent findings in peripheral cholangiocarcinoma located centrally within the liver [20, 30]. In addition, tumor markers such as cancer antigen 19-9 and carcinoembryonic antigen can be clinically useful in establishing a correct differential diagnosis [31, 32]. However, the radiologic diagnosis of small hypovascular nodules in cirrhotic liver on dual-phase helical CT scans remains difficult. We therefore believe that imaging-guided biopsy should be considered before a therapeutic strategy is chosen in the care of patients with liver cirrhosis and small (< 3 cm in diameter) hypoattenuating nodules.
This study had several limitations. First, it was limited by its retrospective nature and by the limited control in selection of the patient population. Second, the number of cholangiocarcinomas was relatively small, and the group of patients was limited to those with a pathologic diagnosis determined at surgery. Third, we used various CT protocols and scanners. Fourth, delayed images (longer than a 10-minute delay) were not included in our routine CT protocol for liver diseases.
The contrast enhancement patterns of intrahepatic, mass-forming cholangiocarcinomas on dual-phase helical CT were found to differ depending on tumor size. Although a large cholangiocarcinoma in a cirrhotic liver can be differentiated from HCC according to the enhancement patterns on CT, a small cholangiocarcinoma can mimic HCC in cirrhotic liver. On the basis of our observations, we believe that in any small hypoattenuating lesion in cirrhotic liver, the diagnosis of cholangiocarcinoma should be seriously considered along with the diagnosis of HCC because of the overlap in imaging findings between the two diseases. Knowledge of the characteristic enhancement patterns of cholangiocarcinoma in cirrhotic liver may aid in the correct diagnosis and choice of appropriate treatment strategy.
Acknowledgments
We thank Bonnie Hami for editorial assistance in the preparation of this
manuscript.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. E. Chung, M.-J. Kim, Y. N. Park, J.-Y. Choi, J. Y. Pyo, Y. C. Kim, H. J. Cho, K. A. Kim, and S. Y. Choi Varying Appearances of Cholangiocarcinoma: Radiologic-Pathologic Correlation1 RadioGraphics, May 1, 2009; 29(3): 683 - 700. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
