Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print October 24, 2002, 10.1148/radiol.2253011346
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2253011346v1
225/3/773    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kim, H.-C.
Right arrow Articles by Choi, B. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kim, H.-C.
Right arrow Articles by Choi, B. I.
(Radiology 2002;225:773-780.)
© RSNA, 2002

Gastrointestinal Imaging

Hepatic Arterial and Portal Venous Phase Helical CT in Patients Treated with Transcatheter Arterial Chemoembolization for Hepatocellular Carcinoma: Added Value of Unenhanced Images1

Hyo-Cheol Kim, MD, Ah Young Kim, MD2, Joon Koo Han, MD, Jin Wook Chung, MD, Jae Young Lee, MD3, Jae Hyung Park, MD and Byung Ihn Choi, MD

1 From the Department of Radiology and Institute of Radiation Medicine, Seoul National University College of Medicine, Clinical Research Institute, Seoul National University Hospital, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea. Received August 8, 2001; revision requested September 28; final revision received April 22, 2002; accepted April 30. Supported in part by 2000 BK21 Project for Medicine, Dentistry, and Pharmacy. Address correspondence to J.K.H. (e-mail: hanjk@radcom.snu.ac.kr).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To evaluate the additional diagnostic value of unenhanced computed tomographic (CT) images in the depiction of viable tumor in patients who were treated with transcatheter arterial chemoembolization (TACE) for hepatocellular carcinoma (HCC) and followed up with biphasic helical CT that included the acquisition of unenhanced images.

MATERIALS AND METHODS: We performed helical CT (with unenhanced, arterial, and portal phases) in 54 patients who had been treated with TACE for HCC. Image analysis was first performed with only those images obtained in the arterial and portal venous phases of helical CT. A second analysis was then performed with unenhanced images, arterial images, and portal venous images that focused on the additional value of unenhanced images. The value of additional unenhanced images was evaluated by means of interobserver agreement ({kappa} statistic) and receiver operating characteristic (ROC) analysis.

RESULTS: The two readers detected 128 and 129 lesions. Unenhanced images were valuable for 32 of 129 lesions (23 patients) for reader 1 and for 29 of 128 lesions (21 patients) for reader 2. Although there was no significant difference between biphasic CT alone and biphasic CT with unenhanced images, results of ROC analysis showed higher diagnostic performance with biphasic CT with unenhanced images than with biphasic CT alone for detecting viable tumor.

CONCLUSION: The study data demonstrate the diagnostic value of unenhanced images interpreted in conjunction with biphasic CT images for follow-up of patients who have previously been treated with TACE for HCC.

© RSNA, 2002

Index terms: Computed tomography (CT), technology, 761.12111, 761.12114 • Liver, CT, 761.12111, 761.12114, 761.12115 • Liver neoplasms, CT, 761.12111, 761.12114, 761.12115


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
With the advent of helical computed tomography (CT) and rapid technical advances, multiphasic helical CT has recently become a popular imaging modality for detecting hypervascular tumors and characterizing liver lesions. In patients with cirrhosis, biphasic CT—that is, CT performed during the hepatic arterial phase (HAP) and portal venous phase (PVP)—has been widely used as the first-line diagnostic modality for detection of hepatocellular carcinoma (HCC), follow-up after local treatment or surgical excision, and assessment of hemodynamic changes in the liver (1). In the past decade, many investigators have reported on the diagnostic efficacy of HAP CT in detecting hypervascular tumors, especially HCC (2–6). However, it has not yet been established whether biphasic CT is the best imaging technique for examining patients with HCC. It is especially unclear whether biphasic CT is the best technique for evaluating the effects of therapy and the possibility of tumor recurrence in patients with HCC who have previously undergone transcatheter arterial chemoembolization (TACE). Magnetic resonance imaging with contrast material enhancement can provide valuable information in patients who have been treated with TACE (7); however, this modality is not as commonly used as CT owing to its increased cost and comparative lack of availability.

In patients treated with TACE for HCC, it is sometimes difficult to monitor viable remaining tumor and marginal and/or remote tumor recurrence with biphasic helical CT because a variety of lesions and phenomena show high attenuation in the HAP that can mimic tumor enhancement. These lesions and phenomena include siderotic nodules, lesions that show faint traces of iodized oil, partial volume averaging effects around masses previously injected with iodized oil as part of TACE, and arterioportal shunts (8). In cases where findings at biphasic helical CT are inconclusive, we believe unenhanced images can be valuable. However, to our knowledge, no study has yet assessed the usefulness of biphasic CT with acquisition of unenhanced images in patients treated with TACE. Therefore, we performed this study to evaluate the additional diagnostic value of unenhanced CT images in the depiction of viable tumor in patients who were treated with TACE for HCC and who were followed up with biphasic helical CT at which unenhanced images were obtained.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Patients
From November 1, 1999, to July 1, 2000, 600 patients who were known or suspected to have HCC were referred for initial or repeat TACE. At a retrospective review (H.C.K.) of the medical records of and radiologic findings in these patients, 54 patients (45 men, nine women; age range, 40–73 years; mean age, 56 years) met the following criteria: (a) a history of previous treatment with TACE for HCC, (b) follow-up with biphasic (ie, HAP and PVP) helical CT at which unenhanced images were also obtained, (c) repeat TACE within 3 weeks after scanning with biphasic helical CT, (d) follow-up with one or more helical CT examinations and/or CT examinations performed to assess the liver for traces of iodized oil after repeat TACE, (e) no previous surgical resection of a mass treated with TACE, and (f) presence of no more than five focal lesions at biphasic helical CT. We excluded 67 patients in whom there were more than five focal lesions because obtaining an accurate lesion count was often impossible and because multiple viable tumors were present in those cases.

Fourteen of the patients included in our study had undergone one previous TACE session, eight had undergone two, nine had undergone three, and 23 had undergone four or more. Our institutional review board does not require its approval or informed consent for review of a patient’s records or images.

On the basis of clinical and radiologic findings, liver cirrhosis associated with viral hepatitis was diagnosed in 51 of 54 patients. The diagnosis of HCC in these patients was rendered on the basis of results of percutaneous needle biopsy (n = 5), results of previous resection of HCC (n = 6), or results (eg, elevated {alpha}-fetoprotein levels and viral markers) of typical clinical or laboratory testing in combination with a typical angiographic appearance and progression of disease on follow-up images (n = 43). It was not possible to obtain lesion-by-lesion histopathologic proof because hepatic resection is not generally performed in patients who are treated with repeat TACE.

CT and TACE Techniques
CT examinations were performed with a Somatom Plus-4 (Siemens Medical Systems, Erlangen, Germany) or a HiSpeed Advantage scanner (GE Medical Systems, Milwaukee, Wis). All patients were asked to drink 200–300 mL of a 2.5% dilution of a sodium amidotrizoate and meglumine amidotrizoate mixture (Gastrografin; Schering, Berlin, Germany) before CT scanning. In all patients, unenhanced scanning was performed through the entire liver with the following parameters: 10-mm collimation, 1:1 table pitch, and 10-mm reconstructions. Each patient received 120 mL of a nonionic contrast material (iopromide, Ultravist 370; Schering Korea, Seoul, Korea) through an 18-gauge angiographic catheter inserted into a forearm vein. The contrast material was injected by means of a Mark V dedicated CT injector (Medrad, Pittsburgh, Pa) at a rate of 3 mL/sec.

In 17 patients, bolus-triggered helical CT with a Somatom Plus-4 scanner was performed with the following parameters: 8-mm collimation and 1:1 table pitch. Contiguous transverse reconstructions were obtained every 8 mm. The scanning delay for the HAP was determined with use of a semiautomatic bolus-tracking program (C.A.R.E. bolus; Siemens Medical Systems), as described in a previous report (9). HAP scanning began 11 seconds after aortic enhancement reached a threshold of 100 HU above the precontrast aortic attenuation value. The delay between contrast material administration and scanning was 65 seconds for the PVP.

In 37 patients, helical CT without bolus tracking was performed with the HiSpeed Advantage scanner, 10-mm collimation, and 1:1 table pitch. Contiguous transverse reconstructions were obtained every 7 mm. The delay between contrast material administration and scanning was 30 seconds for the HAP and 65 seconds for the PVP. Each spiral acquisition through the liver was accomplished during one breath hold.

TACE was performed with a mixture of iodized oil (Lipiodol; Guerbet, Aulnay-sous-Bois, France) and doxorubicin hydrochloride (Adriamycin; Kyowa Hakko Kogyo, Tokyo, Japan) in all patients within 3 weeks after helical CT scanning. All 54 patients were followed up with biphasic helical CT with unenhanced images 25–152 days (mean, 89 days) after TACE. In 29 patients, unenhanced CT scanning through the entire liver was performed 9–30 days (mean, 16 days) after TACE with 10-mm collimation, 1:1 table pitch, and 10-mm reconstructions to assess the liver for traces of iodized oil.

Image Analysis
CT scans were retrospectively and independently reviewed by two radiologists (A.Y.K., J.K.H.). Reviewer 1 had 4 years experience in abdominal imaging and reviewer 2 had 12 years experience. The readers were informed of patient history in that they knew all patients had undergone TACE for HCC, but the readers were blinded to the presence of tumor recurrence in each patient. First, the readers reviewed biphasic images only—the HAP images plus the PVP images—in conjunction with images obtained at helical CT performed before repeat TACE, images obtained at CT performed after TACE to assess the liver for traces of iodized oil, and/or images obtained at angiography. Seven days later, the readers interpreted all three types of images (HAP, PVP, and unenhanced images), focusing on the additional value of unenhanced images in detecting tumor recurrence. The images were presented in random order to each of the readers at each session.

Images were interpreted for the number and relative attenuation of focal lesions; each reader also recorded his or her degree of confidence as to whether a lesion seen on an image represented a viable tumor. The attenuation of each lesion in relation to that of the liver (ie, hypoattenuation, hyperattenuation, isoattenuation, or mixed attenuation) was subjectively assessed and recorded. All images were initially reviewed by using liver window settings (window level, 50–100 HU; window width, 170 HU) at a 2,000 x 2,000 picture archiving and communications system (PACS; Marotech, Seoul, Korea) monitor, and then the window setting in each case was adjusted as needed. Diagnostic confidence for each lesion was scored with a five-point scale (1, not viable tumor; 2, probably not viable tumor; 3, indeterminate; 4, probably viable tumor; 5, definitely viable tumor) in each interpretation session.

For objectivity and reproducibility of the image analysis performed in this study, the criteria for viable tumor and noncancerous lesions were determined. A viable HCC nodule was considered to appear as a hyperattenuating or isoattenuating lesion during the HAP and as a hypoattenuating lesion during the PVP. In addition, the following lesions were regarded as HCC nodules on helical CT images: (a) a nodule of mixed attenuation on HAP images that showed areas of washout on PVP images and (b) a nodule that showed isoattenuation on HAP and PVP images but hypoattenuation on unenhanced images. A lesion showing wedge-shaped hyperattenuation during the HAP that appeared isoattenuating on PVP images and unenhanced images was considered to be a noncancerous lesion.

Reviewers compared biphasic helical CT scans with the images that were used as reference standards in this study (ie, follow-up angiograms, helical CT images, and images obtained at CT performed to assess the liver for traces of iodized oil) to determine the true characteristics of the lesions. The two readers and another reviewer (H.C.K.) defined the findings on the reference standard images in consensus. Because we could not examine every lesion histologically, we presumed that a lesion represented viable tumor if it met the following criteria: (a) it was seen on follow-up angiograms in a location that corresponded to its location on biphasic helical CT images, (b) it showed traces of iodized oil on follow-up CT scans, and (c) it showed an increase in size on follow-up CT scans.

Statistical Analysis
The degree of agreement between the two readers was measured with the {kappa} statistic. The value of additional unenhanced images was evaluated by means of receiver operating characteristic (ROC) analysis. ROC curves were created with a maximum-likelihood curve-fitting algorithm. Statistical analyses were performed with computer software (SPSS, version 10.0; SPSS, Chicago, Ill). A P value of less than .05 was considered to indicate a significant difference.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
A total of 129 lesions were depicted in 54 patients at biphasic helical CT at which unenhanced images were additionally obtained. Reader 1 identified all 129 lesions, while reader 2 identified 128 lesions. Thirteen patients had a single lesion, 14 had two lesions, and 27 had three to five lesions. One hundred twenty-eight lesions were depicted on enhanced images, and one additional lesion (Fig 1) that was not seen on enhanced images was seen only on unenhanced images. Eighty-three lesions contained iodized oil, and 46 newly appeared lesions did not contain iodized oil. Repeat angiography and follow-up CT, which were used as the standards of reference, revealed that 71 lesions were viable tumors, 19 were pseudolesions, and 39 contained iodized oil but were not viable tumors. Forty-six patients had recurrent tumor, and eight patients had no viable tumor.



View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Images in a 70-year-old man who had undergone four treatments with TACE. (a) Transverse HAP CT image and (b) corresponding transverse PVP CT image show no definite lesion in the liver. (c) Corresponding transverse unenhanced CT image shows a 1.5-cm hypoattenuating lesion (arrows). (d) Follow-up angiogram shows nodular tumor staining (arrow) in liver segment 6. Another area of tumor staining (arrowheads) in segment 8 is seen; this finding was also observed on follow-up biphasic helical CT images (not shown).

 


View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Images in a 70-year-old man who had undergone four treatments with TACE. (a) Transverse HAP CT image and (b) corresponding transverse PVP CT image show no definite lesion in the liver. (c) Corresponding transverse unenhanced CT image shows a 1.5-cm hypoattenuating lesion (arrows). (d) Follow-up angiogram shows nodular tumor staining (arrow) in liver segment 6. Another area of tumor staining (arrowheads) in segment 8 is seen; this finding was also observed on follow-up biphasic helical CT images (not shown).

 


View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c. Images in a 70-year-old man who had undergone four treatments with TACE. (a) Transverse HAP CT image and (b) corresponding transverse PVP CT image show no definite lesion in the liver. (c) Corresponding transverse unenhanced CT image shows a 1.5-cm hypoattenuating lesion (arrows). (d) Follow-up angiogram shows nodular tumor staining (arrow) in liver segment 6. Another area of tumor staining (arrowheads) in segment 8 is seen; this finding was also observed on follow-up biphasic helical CT images (not shown).

 


View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d. Images in a 70-year-old man who had undergone four treatments with TACE. (a) Transverse HAP CT image and (b) corresponding transverse PVP CT image show no definite lesion in the liver. (c) Corresponding transverse unenhanced CT image shows a 1.5-cm hypoattenuating lesion (arrows). (d) Follow-up angiogram shows nodular tumor staining (arrow) in liver segment 6. Another area of tumor staining (arrowheads) in segment 8 is seen; this finding was also observed on follow-up biphasic helical CT images (not shown).

 
After additional review of unenhanced images obtained in 24 patients, reader 1 changed the diagnostic confidence level score for 35 of 129 lesions. Assessment of the reclassification of these 35 lesions revealed that review of unenhanced images enabled this reader to be more confident of the characteristics of 32 lesions in 23 patients; this helped reader 1 render the correct diagnosis for 13 lesions that represented viable tumor (Fig 1) and 19 lesions that did not represent viable tumor (Figs 2, 3). Ten of these 32 lesions contained iodized oil (Fig 2), and 22 lesions were newly appeared and did not contain iodized oil (Fig 3). For three of 35 lesions, unenhanced images prompted reader 1 to render a false-negative diagnosis (Fig 4).



View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Images in a 61-year-old man who had undergone eight treatments with TACE. (a) Transverse HAP CT image shows a focal, mildly hyperattenuating area (arrows) between iodized oil-containing nodules previously treated with TACE. (b) Corresponding transverse PVP CT image shows a relatively hypoattenuating lesion (arrows). On the basis of findings of high attenuation on HAP images and low attenuation on PVP images, this lesion was interpreted as a viable tumor. (c) Corresponding transverse unenhanced CT image shows that the hyperattenuating lesion seen in a is mildly hyperattenuating (arrows) in relation to the surrounding hepatic parenchyma; this mild hyperattenuation represents faint traces of iodized oil rather than hypervascular tumor. The strong enhancement of normal hepatic parenchyma during the PVP is why the lesion appears hypoattenuating in b. (d) Follow-up angiogram fails to depict tumor staining. Multiple iodized oil-containing nodules (arrows) that show subtraction artifacts are seen. Multiple areas of faint staining (arrowheads) that were thought to represent arterioportal shunts had a similar appearance on angiograms obtained before repeat TACE (not shown). Also, follow-up helical CT images (not shown) obtained 6 months after repeat TACE revealed no interval change.

 


View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Images in a 61-year-old man who had undergone eight treatments with TACE. (a) Transverse HAP CT image shows a focal, mildly hyperattenuating area (arrows) between iodized oil-containing nodules previously treated with TACE. (b) Corresponding transverse PVP CT image shows a relatively hypoattenuating lesion (arrows). On the basis of findings of high attenuation on HAP images and low attenuation on PVP images, this lesion was interpreted as a viable tumor. (c) Corresponding transverse unenhanced CT image shows that the hyperattenuating lesion seen in a is mildly hyperattenuating (arrows) in relation to the surrounding hepatic parenchyma; this mild hyperattenuation represents faint traces of iodized oil rather than hypervascular tumor. The strong enhancement of normal hepatic parenchyma during the PVP is why the lesion appears hypoattenuating in b. (d) Follow-up angiogram fails to depict tumor staining. Multiple iodized oil-containing nodules (arrows) that show subtraction artifacts are seen. Multiple areas of faint staining (arrowheads) that were thought to represent arterioportal shunts had a similar appearance on angiograms obtained before repeat TACE (not shown). Also, follow-up helical CT images (not shown) obtained 6 months after repeat TACE revealed no interval change.

 


View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c. Images in a 61-year-old man who had undergone eight treatments with TACE. (a) Transverse HAP CT image shows a focal, mildly hyperattenuating area (arrows) between iodized oil-containing nodules previously treated with TACE. (b) Corresponding transverse PVP CT image shows a relatively hypoattenuating lesion (arrows). On the basis of findings of high attenuation on HAP images and low attenuation on PVP images, this lesion was interpreted as a viable tumor. (c) Corresponding transverse unenhanced CT image shows that the hyperattenuating lesion seen in a is mildly hyperattenuating (arrows) in relation to the surrounding hepatic parenchyma; this mild hyperattenuation represents faint traces of iodized oil rather than hypervascular tumor. The strong enhancement of normal hepatic parenchyma during the PVP is why the lesion appears hypoattenuating in b. (d) Follow-up angiogram fails to depict tumor staining. Multiple iodized oil-containing nodules (arrows) that show subtraction artifacts are seen. Multiple areas of faint staining (arrowheads) that were thought to represent arterioportal shunts had a similar appearance on angiograms obtained before repeat TACE (not shown). Also, follow-up helical CT images (not shown) obtained 6 months after repeat TACE revealed no interval change.

 


View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d. Images in a 61-year-old man who had undergone eight treatments with TACE. (a) Transverse HAP CT image shows a focal, mildly hyperattenuating area (arrows) between iodized oil-containing nodules previously treated with TACE. (b) Corresponding transverse PVP CT image shows a relatively hypoattenuating lesion (arrows). On the basis of findings of high attenuation on HAP images and low attenuation on PVP images, this lesion was interpreted as a viable tumor. (c) Corresponding transverse unenhanced CT image shows that the hyperattenuating lesion seen in a is mildly hyperattenuating (arrows) in relation to the surrounding hepatic parenchyma; this mild hyperattenuation represents faint traces of iodized oil rather than hypervascular tumor. The strong enhancement of normal hepatic parenchyma during the PVP is why the lesion appears hypoattenuating in b. (d) Follow-up angiogram fails to depict tumor staining. Multiple iodized oil-containing nodules (arrows) that show subtraction artifacts are seen. Multiple areas of faint staining (arrowheads) that were thought to represent arterioportal shunts had a similar appearance on angiograms obtained before repeat TACE (not shown). Also, follow-up helical CT images (not shown) obtained 6 months after repeat TACE revealed no interval change.

 


View larger version (151K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Images in a 60-year-old man who had undergone three treatments with TACE. (a) Transverse HAP CT image shows an ill-defined hyperattenuating area (arrows). (b) Corresponding transverse PVP CT image shows that the lesion seen in a is isoattenuating; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image also shows that the lesion seen in a is isoattenuating. When all three images were reviewed together, this lesion was interpreted as an arterioportal shunt rather than a hypervascular tumor. (d) Follow-up angiogram shows early opacification of the portal vein (arrows) and focal contrast enhancement of the hepatic parenchyma (arrowheads) in an area that corresponds to the area of the hyperattenuating lesion seen on the HAP CT image. Follow-up CT image (not shown) obtained 6 months after repeat TACE showed no interval change.

 


View larger version (142K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Images in a 60-year-old man who had undergone three treatments with TACE. (a) Transverse HAP CT image shows an ill-defined hyperattenuating area (arrows). (b) Corresponding transverse PVP CT image shows that the lesion seen in a is isoattenuating; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image also shows that the lesion seen in a is isoattenuating. When all three images were reviewed together, this lesion was interpreted as an arterioportal shunt rather than a hypervascular tumor. (d) Follow-up angiogram shows early opacification of the portal vein (arrows) and focal contrast enhancement of the hepatic parenchyma (arrowheads) in an area that corresponds to the area of the hyperattenuating lesion seen on the HAP CT image. Follow-up CT image (not shown) obtained 6 months after repeat TACE showed no interval change.

 


View larger version (142K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c. Images in a 60-year-old man who had undergone three treatments with TACE. (a) Transverse HAP CT image shows an ill-defined hyperattenuating area (arrows). (b) Corresponding transverse PVP CT image shows that the lesion seen in a is isoattenuating; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image also shows that the lesion seen in a is isoattenuating. When all three images were reviewed together, this lesion was interpreted as an arterioportal shunt rather than a hypervascular tumor. (d) Follow-up angiogram shows early opacification of the portal vein (arrows) and focal contrast enhancement of the hepatic parenchyma (arrowheads) in an area that corresponds to the area of the hyperattenuating lesion seen on the HAP CT image. Follow-up CT image (not shown) obtained 6 months after repeat TACE showed no interval change.

 


View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d. Images in a 60-year-old man who had undergone three treatments with TACE. (a) Transverse HAP CT image shows an ill-defined hyperattenuating area (arrows). (b) Corresponding transverse PVP CT image shows that the lesion seen in a is isoattenuating; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image also shows that the lesion seen in a is isoattenuating. When all three images were reviewed together, this lesion was interpreted as an arterioportal shunt rather than a hypervascular tumor. (d) Follow-up angiogram shows early opacification of the portal vein (arrows) and focal contrast enhancement of the hepatic parenchyma (arrowheads) in an area that corresponds to the area of the hyperattenuating lesion seen on the HAP CT image. Follow-up CT image (not shown) obtained 6 months after repeat TACE showed no interval change.

 


View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Images in a 55-year-old man who had undergone two treatments with TACE. (a) Transverse HAP CT image obtained at the lower level of an iodized oil-containing nodule previously treated with TACE shows a hyperattenuating lesion (arrows). (b) Corresponding transverse PVP CT image shows that the lesion (arrows) seen in a is partially "washed out"; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image shows a lesion similar in appearance to that seen in a. When all three images were reviewed together, this lesion was interpreted as probably not being a viable tumor. (d) Follow-up angiogram shows focal nodular tumor staining (arrows) at the bottom and top of the iodized oil-containing nodule (arrowheads) previously treated with TACE. Follow-up scan (not shown) from CT examination performed 2 weeks after repeat TACE to assess the liver for traces of iodized oil showed additional traces of iodized oil at the bottom of the iodized oil-containing nodule previously treated with TACE. The images in this figure illustrate a false-negative result that arose from the evaluation of unenhanced images.

 


View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Images in a 55-year-old man who had undergone two treatments with TACE. (a) Transverse HAP CT image obtained at the lower level of an iodized oil-containing nodule previously treated with TACE shows a hyperattenuating lesion (arrows). (b) Corresponding transverse PVP CT image shows that the lesion (arrows) seen in a is partially "washed out"; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image shows a lesion similar in appearance to that seen in a. When all three images were reviewed together, this lesion was interpreted as probably not being a viable tumor. (d) Follow-up angiogram shows focal nodular tumor staining (arrows) at the bottom and top of the iodized oil-containing nodule (arrowheads) previously treated with TACE. Follow-up scan (not shown) from CT examination performed 2 weeks after repeat TACE to assess the liver for traces of iodized oil showed additional traces of iodized oil at the bottom of the iodized oil-containing nodule previously treated with TACE. The images in this figure illustrate a false-negative result that arose from the evaluation of unenhanced images.

 


View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c. Images in a 55-year-old man who had undergone two treatments with TACE. (a) Transverse HAP CT image obtained at the lower level of an iodized oil-containing nodule previously treated with TACE shows a hyperattenuating lesion (arrows). (b) Corresponding transverse PVP CT image shows that the lesion (arrows) seen in a is partially "washed out"; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image shows a lesion similar in appearance to that seen in a. When all three images were reviewed together, this lesion was interpreted as probably not being a viable tumor. (d) Follow-up angiogram shows focal nodular tumor staining (arrows) at the bottom and top of the iodized oil-containing nodule (arrowheads) previously treated with TACE. Follow-up scan (not shown) from CT examination performed 2 weeks after repeat TACE to assess the liver for traces of iodized oil showed additional traces of iodized oil at the bottom of the iodized oil-containing nodule previously treated with TACE. The images in this figure illustrate a false-negative result that arose from the evaluation of unenhanced images.

 


View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4d. Images in a 55-year-old man who had undergone two treatments with TACE. (a) Transverse HAP CT image obtained at the lower level of an iodized oil-containing nodule previously treated with TACE shows a hyperattenuating lesion (arrows). (b) Corresponding transverse PVP CT image shows that the lesion (arrows) seen in a is partially "washed out"; this lesion was initially interpreted as probably being a viable tumor. (c) Corresponding transverse unenhanced CT image shows a lesion similar in appearance to that seen in a. When all three images were reviewed together, this lesion was interpreted as probably not being a viable tumor. (d) Follow-up angiogram shows focal nodular tumor staining (arrows) at the bottom and top of the iodized oil-containing nodule (arrowheads) previously treated with TACE. Follow-up scan (not shown) from CT examination performed 2 weeks after repeat TACE to assess the liver for traces of iodized oil showed additional traces of iodized oil at the bottom of the iodized oil-containing nodule previously treated with TACE. The images in this figure illustrate a false-negative result that arose from the evaluation of unenhanced images.

 
After additional review of unenhanced images obtained in 22 patients, reader 2 changed the diagnostic confidence level score for 31 of 128 lesions. Assessment of the reclassification of these 31 lesions revealed that review of unenhanced images enabled reader 2 to be more confident of the characteristics of 29 lesions in 21 patients. Twenty of these 29 lesions proved to be noncancerous pseudolesions, and 11 of these 29 lesions contained iodized oil. For two of 31 lesions, unenhanced images prompted reader 2 to render false-negative diagnoses. Neither reader rendered a false-positive diagnosis.

In three of eight patients who were found to have no viable tumor at follow-up angiography and CT, the two readers initially suspected that four lesions seen on enhanced images represented viable tumors, but concluded that no viable tumors were probably present after they reviewed unenhanced images.

Interobserver agreement ({kappa} statistic) for the presence of viable tumor was 0.876 for the review of biphasic CT images (ie, HAP images plus PVP images) and 0.884 when all three types of images (HAP, PVP, and unenhanced images) were reviewed. Interobserver agreement for the biphasic helical CT images was considered to be very good. The area under the ROC curve for reader 1 was 0.934 for the biphasic CT images and 0.983 for all three types of images (Fig 5a). The area under the curve for reader 2 was 0.955 for the biphasic CT images and 0.985 for all three types of images (Fig 5b). At the 95% confidence level, the CIs for reader 1 were 0.891, 0.97 with the biphasic images and 0.961, 1.00 with all three types of images. The 95% CIs of reader 2 were 0.926, 0.985 with the biphasic images and 0.965, 1.00 with all three types of images.



View larger version (33K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a. Graphs of ROC analysis results clearly show additional value of unenhanced images to both readers in detecting viable tumor. Dotted line = biphasic images, solid line = biphasic images and unenhanced images. (a) The area under the curve for reader 1 was 0.934 with the biphasic images and 0.983 with all three types of images. (b) The area under the curve for reader 2 was 0.955 with the biphasic images and 0.985 with all three types of images.

 


View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b. Graphs of ROC analysis results clearly show additional value of unenhanced images to both readers in detecting viable tumor. Dotted line = biphasic images, solid line = biphasic images and unenhanced images. (a) The area under the curve for reader 1 was 0.934 with the biphasic images and 0.983 with all three types of images. (b) The area under the curve for reader 2 was 0.955 with the biphasic images and 0.985 with all three types of images.

 
Comparison of the areas under the ROC curves for the detection of viable tumor demonstrated a nonsignificant difference (P > .05) between the biphasic CT images and all three types of images for the two readers. Although there was no significant difference between the biphasic CT images and all three types of images, addition of unenhanced images increased diagnostic confidence in detecting viable tumor in patients who had undergone TACE.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
REFERENCES
 
Although biphasic (HAP plus PVP) helical CT is usually used to monitor the therapeutic effect of TACE, there is considerable difficulty in accurately detecting residual viable tumor or recurrent tumor with this modality. The high attenuation of iodized oil makes it difficult to differentiate areas of true enhancement from areas of retained iodized oil.

The role of unenhanced CT of the liver remains controversial. Before the advent of helical CT, some researchers recommended that both unenhanced and contrast material–enhanced CT images be obtained in patients suspected of having HCC or other hypervascular liver tumors (10,11), while others argued that the yield obtained from unenhanced scanning does not warrant the added time, cost, and radiation exposure it entails (12,13). In this era of biphasic helical CT, Oliver et al (14) have reported that the use of unenhanced images together with PVP images enables detection of significantly more hypervascular liver metastases than does the use of HAP images together with PVP images. However, Oliver et al (15) have also reported that in patients with HCC, the combination of HAP and PVP images revealed significantly more HCC than did the combination of unenhanced and PVP images (15). Other investigators have questioned whether unenhanced CT images can depict additional foci of HCC or hypervascular metastatic disease (16,17).

Although it is possible that the increasing use of biphasic helical CT will lead to an improvement in the detection of HCC (2–6), not all HCC nodules are hypervascular and not all hypervascular HCC nodules will appear as hyperattenuating lesions within the liver during the HAP. Because of the diverse vascular nature of these lesions, some will enhance rapidly, achieving the same degree of enhancement and attenuation as the background liver parenchyma, and hence they will not be visualized (15). In addition, iodized oil–containing nodules can simulate hypervascular tumors on HAP images obtained in patients who have been treated with TACE. Therefore, the question remains: Does unenhanced CT imaging still have a prominent role in the detection of HCC in patients treated with TACE in this era of biphasic helical contrast-enhanced CT? Is it possible to avoid the potential pitfalls of isoattenuating HCC nodules and iodized oil–containing nodules that mimic hypervascular tumors by routinely obtaining unenhanced CT images of the liver before contrast-enhanced CT imaging is performed?

Clearly, unenhanced images are superior to HAP images or PVP images for detecting calcified lesions. Iodized oil behaves like calcification, so unenhanced images have added value in detecting viable tumor. In our study, 10 of 32 lesions for which unenhanced images had additional value for reader 1 and 11 of 29 lesions for which unenhanced images had additional value for reader 2 contained iodized oil. Iodized oil–containing nodules usually appeared as very highly attenuating round nodules in which the unenhanced images usually did not confer added value. On the other hand, unenhanced images were frequently valuable in evaluating nodules in which the iodized oil was partially washed out so the nodule appeared to have a focal defect within it and nodules in which the iodized oil grew pale and the nodule appeared to have fuzzy outlines. Unenhanced images were also valuable when assessment of residual or recurring tumor on enhanced-phase images was complicated by a partial volume averaging effect caused by an iodized oil–containing nodule.

A small transient hepatic attenuation difference is most often caused by idiopathic obstruction of a small peripheral portal branch or by arterioportal shunts. Areas of portal branch obstruction are almost always wedge shaped and peripheral in location. Still, it can occasionally be difficult to differentiate portal branch obstruction from a hypervascular neoplasm. Further clues to the correct diagnosis of such obstructions, beyond their location and shape, include straight-line margins and the presence of normal vessels coursing through the hypervascular area (1). Another cause of the transient hepatic attenuation difference effect is arterioportal shunts. These shunts may be due to previous TACE, although other causes of such shunts exist, including HCC and cirrhosis (1,7). The result of redistributing arterial flow and contrast material into a focal region of portal venous flow is the focal enhancement of the region of the liver receiving the enhanced portal flow, while the rest of the liver is predominantly unenhanced during the HAP. Hence, the transient hepatic attenuation difference effect results in the appearance of a hypervascular lesion that is sometimes indistinguishable from a hypervascular tumor.

If the area of high attenuation on HAP images shows isoattenuation on unenhanced images, reviewers are typically more confident that it represents a noncancerous pseudolesion rather than a malignant condition. In this study, 19 of 32 lesions for which unenhanced images had additional value for reader 1 and 20 of 29 lesions for which unenhanced images had additional value for reader 2 were diagnosed as noncancerous pseudolesions on the basis of findings at follow-up imaging. Most of these pseudolesions may have been tiny arterioportal shunts, although early opacification of a branch of the portal vein was not commonly demonstrated at angiography because of limited resolution and presence of overlapping liver parenchyma or bowel.

The efficacy and tolerability of TACE can be improved if it is used selectively and is repeated only when necessary and according to the results of follow-up CT (18). In the present study, in three of eight patients who had no viable tumor at follow-up angiography and CT, unenhanced images increased reviewers’ confidence that no viable tumor was present. Biphasic helical CT with additional unenhanced images may help patients like these avoid unnecessary TACE.

The possible causes of the false-negative interpretations of unenhanced images encountered in our study are thought to be the following: (a) Corresponding enhanced and unenhanced images were not obtained at identical levels, due to patient breathing, and (b) partial volume averaging effects make it difficult to detect small viable tumors recurring at the top or bottom of iodized oil–containing masses. It is thought that these obstacles might be avoided at three-dimensional breath-hold multi–detector row helical CT of the liver with use of multiplanar reformation.

Our study had certain limitations. First, our study was a retrospective analysis of data acquired with various CT scanners. In addition, constant time delay was used in 37 patients and bolus-triggered imaging was used in 17 patients for the HAP. However, in our experience, the bolus-triggered HAP began within a few seconds of our constant time delays in all patients, except a few with severe cardiac problems. Second, a detailed lesion-by-lesion histopathologic analysis was not possible. However, we confirmed viable tumors thought to be detected at biphasic helical CT with follow-up angiography, CT performed to assess the liver for traces of iodized oil, and helical CT. Third, we excluded patients who had more than five focal lesions, and this may have contributed to high interobserver agreement. However, we believe our results imply that the ability to detect viable tumor with biphasic helical CT with additional unenhanced imaging would be similarly increased in such patients. Fourth, section thickness was 7 mm or 8 mm for enhanced images and 10 mm for unenhanced images. It would have been ideal for each corresponding image to have been obtained at an identical body level and at an identical section thickness, but the patients took a breath between each phase of scanning, which made it impossible for each corresponding image to have an identical body level. Moreover, the different section thicknesses sometimes rendered the corresponding unenhanced images incomparable with the enhanced images. Because the unenhanced images had additional value in our study in spite of their different section thicknesses, unenhanced and enhanced images obtained with identical section thicknesses might yield even greater value in the diagnosis of viable tumor.

In conclusion, unenhanced images can provide additional diagnostic value to biphasic helical CT images in patients who have previously been treated with TACE for HCC by facilitating differentiation between true lesions and hyperattenuating pseudolesions that simulate viable tumor. The addition of unenhanced imaging offers increased ability to detect HCC, as well as increased diagnostic confidence in the assessment of viable tumor, compared with biphasic helical CT alone. Therefore, we recommend the use of biphasic helical CT at which unenhanced images are additionally obtained for the evaluation of patients treated with TACE for HCC who are suspected of having recurring tumor.


   FOOTNOTES
 
2 Current address: Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea. Back

3 Current address: Department of Radiology, Hallym University Sacred Heart Hospital, Anyang, Korea. Back

Abbreviations: HAP = hepatic arterial phase, HCC = hepatocellular carcinoma, PVP = portal venous phase, ROC = receiver operating characteristic, TACE = transcatheter arterial chemoembolization

Author contributions: Guarantor of integrity of entire study, B.I.C.; study concepts, J.K.H.; study design, J.K.H., A.Y.K.; literature research, H.C.K., J.Y.L.; clinical studies, J.K.H., J.W.C., J.H.P.; data acquisition, H.C.K.; data analysis/interpretation, A.Y.K., J.K.H.; statistical analysis, A.Y.K.; manuscript preparation, H.C.K.; manuscript definition of intellectual content, H.C.K., A.Y.K.; manuscript editing, A.Y.K., J.K.H.; manuscript revision/review, J.K.H.; manuscript final version approval, B.I.C.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 

  1. Oliver JH, III, Baron RL. Helical biphasic contrast-enhanced CT of the liver: techniques, indications, interpretation, pitfalls. Radiology 1996; 201:1-14.[Abstract/Free Full Text]
  2. Bonaldi VM, Bret PM, Reinhold C, Atri M. Helical CT of the liver: value of an early arterial phase. Radiology 1995; 197:357-363.[Abstract/Free Full Text]
  3. Baron RL, Oliver JH, III, Dodd GD, III, Nalesnik M, Holbert BL, Carr B. Hepatocellular carcinoma: evaluation with biphasic, contrast-enhanced, helical CT. Radiology 1996; 199:505-511.[Abstract/Free Full Text]
  4. Hwang GJ, Kim MJ, Yoo HS, Lee JT. Nodular hepatocellular carcinomas: detection with arterial-, portal-, and delayed-phase images at spiral CT. Radiology 1997; 202:383-388.[Abstract/Free Full Text]
  5. Kanematsu M, Oliver JH, III, Carr B, Baron RL. Hepatocellular carcinoma: the role of helical biphasic contrast-enhanced CT versus CT during arterial portography. Radiology 1997; 205:75-80.[Abstract/Free Full Text]
  6. Choi BI, Lee HJ, Han JK, Choi DS, Seo JB, Han MC. Detection of hypervascular nodular hepatocellular carcinomas: value of triphasic helical CT compared with iodized-oil CT. AJR Am J Roentgenol 1997; 168:219-224.[Abstract/Free Full Text]
  7. Kubota K, Hisa N, Nishikawa T, et al. Evaluation of hepatocellular carcinoma after treatment with transcatheter arterial chemoembolization: comparison of Lipiodol-CT, power Doppler sonography, and dynamic MRI. Abdom Imaging 2001; 26:184-190.[CrossRef][Medline]
  8. Kim TK, Choi BI, Han JK, Chung JW, Park JH, Han MC. Nontumorous arterioportal shunt mimicking hypervascular tumor in cirrhotic liver: two-phase spiral CT findings. Radiology 1998; 208:597-603.[Abstract/Free Full Text]
  9. Kopka L, Funke M, Fischer U, Vosshenrich R, Oestmann JW, Grabbe E. Parenchymal liver enhancement with bolus-triggered helical CT: preliminary clinical results. Radiology 1995; 195:282-284.[Abstract/Free Full Text]
  10. Bressler EL, Alpern MB, Glazer GM, Francis IR, Ensminger WD. Hypervascular hepatic metastases: CT evaluation. Radiology 1987; 162:49-51.[Abstract/Free Full Text]
  11. DuBrow RA, David CL, Libshitz HI, Lorigan JG. Detection of hepatic metastases in breast cancer: the role of nonenhanced and enhanced CT scanning. J Comput Assist Tomogr 1990; 14:366-369.[Medline]
  12. Patten RM, Byun JY, Freeny PC. CT of hypervascular hepatic tumors: are unenhanced scans necessary for diagnosis? AJR Am J Roentgenol 1993; 161:979-984.[Abstract/Free Full Text]
  13. Chomyn JJ, Stamm ER, Thickman D. CT of melanoma liver metastases: is the examination without contrast media superfluous? J Comput Assist Tomogr 1992; 16:568-571.[Medline]
  14. Oliver JH, III, Baron RL, Federle MP, Jones BC, Sheng R. Hypervascular liver metastases: do unenhanced and hepatic arterial phase CT images affect tumor detection? Radiology 1997; 205:709-715.[Abstract/Free Full Text]
  15. Oliver JH, III, Baron RL, Federle MP, Rockette HP. Detecting hepatocellular carcinoma: the relative added value of using noncontrast and/or arterial phase imaging in conjunction with conventional portal venous phase contrast CT. AJR Am J Roentgenol 1996; 167:71-77.[Abstract/Free Full Text]
  16. Miller FH, Butler RS, Hoff FL, Fitzgerald SW, Nemcek AA, Gore RM. Using triphasic helical CT to detect focal hepatic lesions in patients with neoplasms. AJR Am J Roentgenol 1998; 171:643-649.[Abstract/Free Full Text]
  17. Sheafor DH, Frederick MG, Paulson EK, Keogan MT, DeLong DM, Nelson RC. Comparison of unenhanced, hepatic arterial-dominant, and portal venous-dominant phase helical CT for the detection of liver metastases in women with breast carcinoma. AJR Am J Roentgenol 1999; 172:961-968.[Abstract/Free Full Text]
  18. Ernst O, Sergent G, Mizrahi D, Delemazure O, Paris JC, L’Herminé C. Treatment of hepatocellular carcinoma by transcatheter arterial chemoembolization: comparison of planned periodic chemoembolization and chemoembolization based on tumor response. AJR Am J Roentgenol 1999; 172:59-64.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
S. H. Kim, J. M. Lee, Y. J. Kim, J. Y. Choi, G. H. Kim, H. Y. Lee, and B. I. Choi
Detection of Hepatocellular Carcinoma on CT in Liver Transplant Candidates: Comparison of PACS Tile and Multisynchronized Stack Modes
Am. J. Roentgenol., May 1, 2007; 188(5): 1337 - 1342.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
P. V. Pandharipande, G. A. Krinsky, H. Rusinek, and V. S. Lee
Perfusion Imaging of the Liver: Current Challenges and Future Goals
Radiology, March 1, 2005; 234(3): 661 - 673.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
R. Iannaccone, A. Laghi, C. Catalano, P. Rossi, F. Mangiapane, T. Murakami, M. Hori, F. Piacentini, I. Nofroni, and R. Passariello
Hepatocellular Carcinoma: Role of Unenhanced and Delayed Phase Multi-Detector Row Helical CT in Patients with Cirrhosis
Radiology, February 1, 2005; 234(2): 460 - 467.
[Abstract] [Full Text] [PDF]

Home page
 RadioGraphicsHome page
A. E. Flanders, R. H. Wiggins III, and M. E. Gozum
Handheld Computers in Radiology
RadioGraphics, July 1, 2003; 23(4): 1035 - 1047.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2253011346v1
225/3/773    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kim, H.-C.
Right arrow Articles by Choi, B. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kim, H.-C.
Right arrow Articles by Choi, B. I.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS