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


     

Published online before print September 28, 2005, 10.1148/radiol.2372041183
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2372041183v1
237/2/520    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 Lutz, A. M.
Right arrow Articles by Weishaupt, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lutz, A. M.
Right arrow Articles by Weishaupt, D.
(Radiology 2005;237:520-528.)
© RSNA, 2005

Gastrointestinal Imaging

Hepatocellular Carcinoma in Cirrhosis: Enhancement Patterns at Dynamic Gadolinium- and Superparamagnetic Iron Oxide–enhanced T1-weighted MR Imaging1

Amelie M. Lutz, MD, Jürgen K. Willmann, MD, Kerstin Goepfert, RT, Borut Marincek, MD and Dominik Weishaupt, MD

1 From the Institute of Diagnostic Radiology, University Hospital Zurich, Raemistrasse 100, 8091 Zurich, Switzerland. Received July 6, 2004; revision requested September 14; revision received December 6; accepted January 18, 2005. Address correspondence to D.W. (e-mail: dominik.weishaupt{at}dmr.usz.ch).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
PURPOSE: To prospectively compare intraindividual differences in enhancement patterns between gadolinium- and superparamagnetic iron oxide (SPIO)-enhanced magnetic resonance (MR) imaging in patients with histologically proved hepatocellular carcinoma (HCC).

MATERIALS AND METHODS: Institutional review board approval and informed consent were obtained. Twenty-two patients (18 men, four women; mean age, 58.9 years) with 36 pathologically proved HCC lesions underwent contrast material–enhanced dynamic T1-weighted gradient-echo MR imaging twice. Gadopentetate dimeglumine was used at the first session. After a mean interval of 5 days, a second session was performed with a bolus-injectable SPIO agent, ferucarbotran. Qualitative analysis of contrast enhancement patterns with each agent during hepatic arterial, portal venous, and equilibrium phases was performed by two readers who classified lesions as isointense, hypointense, or hyperintense compared with surrounding liver parenchyma and searched for presence of hyperintense peritumoral ring enhancement. Results of signal intensity analysis during different vascular phases at both sessions were compared by using the McNemar test, and {kappa} statistic was used to evaluate agreement between signal intensity and enhancement pattern of lesions during different vascular phases.

RESULTS: On gadolinium-enhanced hepatic arterial phase images, HCC lesions (n = 36) were hyperintense in 21 (58%) cases, hypointense in 10 (28%), and isointense in five (14%). On ferucarbotran-enhanced hepatic arterial phase images, HCC lesions were isointense in 18 (50%) cases, hypointense in 11 (31%), and hyperintense in seven (19%). On gadolinium-enhanced portal venous and equilibrium phase images, respectively, HCC lesions were hypointense in 17 (47%) and 21 (58%) cases, hyperintense in 10 (28%) cases and one (3%) case, and isointense in nine (25%) and 14 (39%) cases. On ferucarbotran-enhanced portal venous and equilibrium phase images, respectively, HCC lesions were hypointense in 15 (42%) and 11 (31%) cases, hyperintense in three (8%) and three (8%) cases, and isointense in 18 (50%) and 22 (61%) cases.

CONCLUSION: For HCC, contrast enhancement pattern on T1-weighted gradient-echo MR images shows marked variability with gadolinium or SPIO contrast agents.

© RSNA, 2005


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Hepatocellular carcinoma (HCC) is the most common primary liver tumor, and viral hepatitis and alcoholic cirrhosis are the leading risk factors (1). The prevalence of HCC shows a distinct geographic variation, ranging up to 100 diseased patients per a population of 100 000 in some parts of Asia and Africa (1,2). Although the prevalence of HCC in Europe and the United States is reported to be lower, it is considered an increasing health care problem in these places.

Over the past years, the 5-year survival rate for patients with HCC has improved, even for those with advanced tumor stages. With advanced treatment strategies, including partial hepatectomy (3), liver transplantation (4), radiofrequency ablation (5,6), percutaneous ethanol injections (7), transarterial chemoembolization (8), or often a combination of the different methods, overall 5-year survival rates of up to 70% can be achieved in patients, with adequate preservation of hepatic function (9).

Imaging plays a key role not only in the pretherapeutic assessment of HCC but also in the monitoring of the tumor after therapy. Over the past decade, magnetic resonance (MR) imaging has gained a major role in the diagnostic work-up of HCC, in particular in patients with liver cirrhosis. Although there is general agreement that contrast material–enhanced MR imaging is superior to unenhanced MR imaging in the detection of HCC in the cirrhotic liver, it is still controversial whether the use of extracellular contrast agents, superparamagnetic iron oxide (SPIO) particles, or the combination of both is the most sensitive for depiction of HCC (10–13).

So far, most studies that have been performed with SPIO as a contrast agent for the assessment of HCC were performed with ferumoxides (Feridex, Advanced Magnetics, Cambridge, Mass; Endorem, Guerbet, Aulnay sous Bois, France). Although ferumoxides shorten T1 relaxation time (14,15), most of these studies were performed to evaluate the T2 and T2* properties of this contrast agent. Since the bolus injection of ferumoxides is not recommended because of possible side effects, dynamic contrast-enhanced imaging has not been possible so far. Recently, ferucarbotran (Resovist; Schering, Berlin, Germany) became available as a new SPIO agent for liver imaging in most European countries, as well as in some countries in Asia. Ferucarbotran is an SPIO agent that can be injected as a bolus (16), at a rate of 2 mL/sec for example, which enables dynamic MR imaging to be performed during different vascular phases as we are accustomed to in liver imaging with extracellular contrast agents. In the accumulation phase, when the SPIO particles are taken up by the Kupffer cells of normal liver parenchyma or by Kupffer cells located in benign liver lesions, T2 and T2* effects and, less frequently, T1 effects, are used in lesion detection and characterization (17,18).

With use of conventional extracellular contrast agents, such as gadolinium chelates, analysis of enhancement patterns at T1-weighted dynamic imaging during the different vascular phases is an important tool in the detection and characterization of focal liver lesions, especially of small liver lesions in the cirrhotic liver, because many of these lesions are occult at unenhanced MR imaging. Through our experience with ferucarbotran, we learned that the well-known enhancement patterns of extracellular contrast agents do not readily correlate with those at SPIO-enhanced T1-weighted dynamic MR imaging. To our knowledge, no study has been performed to examine the typical enhancement patterns of HCC at SPIO-enhanced dynamic T1-weighted MR imaging. Therefore, the purpose of this study was to prospectively compare the intraindividual differences in enhancement patterns between gadolinium- and SPIO-enhanced MR imaging in patients with histologically proved HCC.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Patients
During a 5-month period, 22 consecutive patients known to have or suspected of having HCC and underlying liver cirrhosis (18 men and four women; mean age, 58.9 years; age range, 34–80 years) were prospectively included in this study. Patients were included in the study only if they had no prior therapy for HCC, including liver surgery, liver transplantation, radiofrequency ablation, percutaneous ethanol injection, or transarterial chemoembolization. None of the patients underwent biopsy before MR imaging, and none of the patients had undergone placement of a transjugular intrahepatic portosystemic stent. All patients were referred to our institution for hepatic MR imaging for clinical purposes from our Department of Gastroenterology and the Department of Visceral and Transplantation Surgery. The study was approved by our hospital's institutional review board, and written informed consent was obtained from all patients.

The underlying cause of liver cirrhosis was alcohol abuse (seven patients), hepatitis C (seven patients), alcohol abuse combined with hepatitis C (three patients), hepatitis B (four patients), or primary biliary cirrhosis (one patient). Liver cirrhosis was histologically confirmed in all patients. Cirrhosis was clinically classified as Child-Pugh class A in eight patients, Child-Pugh class B in 12 patients, and Child-Pugh class C in two patients.

The histologic specimens were obtained by means of percutaneous biopsy in 13 patients, by means of intraoperative biopsy in six patients, and following liver transplantation in three patients. Histopathologic examination of lesion specimens obtained at percutaneous or intraoperative biopsy (wedge resection) or following liver transplantation revealed a total of 36 HCC lesions in 22 patients (14 patients had one lesion each, two patients had two lesions each, and six patients had three lesions each). The mean interval between MR imaging and transplantation was 295 days (range, 22–812 days). The histologic differentiation was good in 21 lesions, moderate in seven lesions, and poor in eight lesions.

MR Imaging
All patients underwent two MR examinations of the liver, which were separated by a mean interval of 5 days (range, 1–24 days). The precontrast MR sequences were performed similarly at both imaging sessions. For the dynamic contrast-enhanced T1-weighted sequences, gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) was used in the first imaging session, and ferucarbotran (Resovist; Schering) was used in the second imaging session (see imaging protocol). To avoid interference due to the prolonged T2 effect of ferucarbotran, gadopentetate dimeglumine was used as the contrast agent in the first imaging session in all patients.

All MR imaging examinations were performed by using a 1.5-T system (Signa CV/i; GE Medical Systems, Milwaukee, Wis). An anteroposterior phased-array surface coil (torso-array coil) for signal reception was placed around the patient and covered the entire liver in all patients. A respiratory belt was placed around the patient's upper abdomen for the acquisition of respiratory-triggered sequences.

The imaging protocol at both sessions included a transverse unenhanced respiratory-triggered T2-weighted fast spin-echo sequence with fat suppression (repetition time msec/echo time msec, 3500–5714/99–106; echo train length, 12; two signals acquired; field of view, 32–35 x 22–24 cm; matrix, 256 x 256; section thickness, 6.0–8.0 mm; intersection gap, 2.0 mm), as well as transverse unenhanced T1-weighted fast multiplanar spoiled gradient-recalled echo (GRE) sequences both in phase and out of phase (150–200/4.2–4.4 for in-phase and 120–180/1.5–2.2 for out-of-phase acquisition; flip angle, 60°; one signal acquired; field of view, 32–35 x 22–24 cm; matrix, 256 x 192; section thickness, 6.0–8.0 mm with no intersection gap). In addition, a T2-weighted single-shot fast spin-echo sequence (62 400/96; 0.5 signal acquired; field of view, 36–40 x 36–40 cm; matrix, 256 x 24; section thickness, 7.0 mm with no intersection gap) was performed in the coronal plane prior to intravenous contrast material administration.

Subsequently, transverse dynamic frequency-selective fat saturated T1-weighted two-dimensional fast spoiled GRE MR sequences were performed before and after administration of the contrast agent (first imaging session, gadopentetate dimeglumine; second imaging session, ferucarbotran) in the hepatic arterial phase (20 seconds after contrast agent administration), the portal venous phase (60 seconds after contrast agent administration), and the equilibrium phase (240 seconds after contrast agent administration). The parameters for the two-dimensional GRE sequence were as follows: 150–170/1.4; flip angle, 60°; one signal acquired; field of view, 32–35 x 22–24 cm; matrix, 256 x 192; and section thickness, 6.0–8.0 mm with no intersection gap.

Gadopentetate dimeglumine was administered as an intravenous bolus at a dose of 0.2 mL per kilogram of body weight (corresponding to 0.1 mmol per kilogram of body weight) with a flow rate of 2 mL/sec, which was followed with a 20-mL saline flush at the same flow rate by using a power injector (Spectris; Medrad, Indianola, Pa). Ferucarbotran was administered as an intravenous bolus at a dose of 1.4 mL (corresponding to 0.7 mmol iron) for the contrast-enhanced sequences at a flow rate of approximately 2 mL/sec through a system with a 5-µm in-line filter (located between the infusion tube and the intravenous cannula), which was followed with a 20-mL saline flush at the same flow rate by using the same power injector as in the first session. Since the volume of ferucarbotran was relatively small, the contrast agent was prefilled to the distal part of the intravenous line with the tip of the contrast agent bolus located immediately before the in-line filter and the peripheral venous access device. This was possible because of the brownish color of ferucarbotran. Saline was then injected to flush the contrast agent into the patient's venous system.

For ferucarbotran-enhanced MR imaging, at least 10 minutes after ferucarbotran injection, additional T2-weighted fast spin-echo sequences were performed in the transverse plane with fat suppression (3500–5714/99–106; echo train length, 12; two signals acquired; field of view, 32–35 x 22–24 cm; matrix, 256 x 256; section thickness, 6.0–8.0 mm with no intersection gap), as well as T2*-weighted GRE sequences (300/7.2; flip angle, 20°; 0.75 signal acquired; field of view, 32–35 x 22–24 cm; matrix, 256 x 192; section thickness, 6.0–8.0 mm with no intersection gap).

Image Analysis
Image analysis was performed electronically at a dedicated workstation (Advantage Windows Workstation; GE Medical Systems Europe, Buc, France). Two observers (D.W. and J.K.W., 10 and 3 years of experience in liver MR imaging, respectively) independently reviewed all MR images separately in two different reading sessions that were separated by a 4-week interval. The observers analyzed the gadolinium-enhanced MR image sets at the first reading session and the ferucarbotran-enhanced MR image sets at the second reading session. The order of the MR image sets was random in both reading sessions. Both observers were blinded to any clinical information and to which contrast agent was used. During each reading session, the precontrast and postcontrast T1-weighted GRE images were available. The other sequences were not available for readout. For the ferucarbotran-enhanced image sets, the postcontrast T2- and T2*-weighted images (in the accumulation phase) were also not available for readout. Differences among the observers were resolved by means of consensus conference. Overall, a consensus reading was necessary in 17 of 252 instances. To ensure that both observers assessed the same liver lesion and that only the histopathologically confirmed liver lesions were analyzed, the study coordinator (A.M.L.) designated the liver lesions to be analyzed on a segmental liver map. This map was distributed to both observers prior to each reading session. The study coordinator also measured the maximum diameter of the lesions on images obtained with three different sequences (precontrast T2-weighted, precontrast T1-weighted, and postcontrast T1-weighted during the hepatic arterial phase) on the same workstation by using electronic calipers and calculating the mean size for each lesion.

For qualitative analysis, the observers were asked to evaluate the signal intensity characteristics and enhancement patterns of the lesions assigned by the study coordinator on precontrast T1-weighted as well as on postcontrast dynamic T1-weighted GRE images. The signal intensity characteristics of the lesion compared with those of the surrounding liver parenchyma were classified on the precontrast T1-weighted images as well as on images obtained from each phase of the postcontrast T1-weighted GRE sequence separately by using the following classifications: isointense, hypointense, and hyperintense. When lesions showed an inhomogeneous enhancement pattern on any of the postcontrast images, they were categorized as isointense, hyperintense, or hypointense lesions according to the signal intensity characteristics of the predominant parts of the lesions. In addition, the presence of hyperintense peritumoral ring enhancement during any vascular phase was noted.

Statistical Analysis
The results of the signal intensity analysis during the different vascular phases at both imaging sessions were compared by using the McNemar test. P ≤ .05 was considered to indicate statistical significance. Because of the small number of patients with multiple lesions (eight of 22), statistical analysis considering data clustering was not performed.

{kappa} statistic was performed to evaluate the agreement between the signal intensities and the enhancement pattern of the lesions during the different vascular phases with the contrast agents. Results were expressed as {kappa} values and could be classified according to the scale of Landis and Koch (19) as follows: poor, <0; slight, 0.0–0.20; fair, 0.21–0.40; moderate, 0.41–0.60; good, 0.61–0.80; and excellent, 0.81–1.00).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Lesion Size
All 36 lesions were visible after administration of either contrast agent during at least one vascular phase on the dynamic T1-weighted GRE images. The mean diameter of the lesions at MR imaging was 3.1 cm (range, 1.3–6.9 cm). Ten lesions were 2 cm or smaller, seven lesions were 2–3 cm, and 19 lesions were larger than 3 cm. There was no statistically significant difference with regard to the diameter of the lesions between the gadolinium-enhanced and the ferucarbotran-enhanced MR data sets.

Qualitative Analysis
The results of the qualitative image analysis are displayed in the Table.


View this table:
[in this window]
[in a new window]

  		Qualitative Analysis of Enhancement Patterns in 36 Proved HCC Lesions at MR Imaging

 
On precontrast T1-weighted GRE images, the majority of lesions were hypointense (30 of 36 lesions, 83%); only a few lesions were isointense (four of 36 lesions, 11%) or hyperintense (two of 36 lesions, 6%).

On gadolinium-enhanced T1-weighted GRE images obtained during the hepatic arterial phase, most HCC lesions (21 of 36, 58%) were hyperintense relative to the surrounding liver parenchyma (Figs 1–3). In contrast, on ferucarbotran-enhanced T1-weighted GRE images obtained during the hepatic arterial phase, the majority of the lesions were either isointense (18 of 36, 50%) (Fig 1) or hypointense (11 of 36, 31%) (Figs 2, 3), and the minority were hyperintense (seven of 36, 19%).



View larger version (165K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (166K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (166K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (161K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1e. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1f. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1g. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VIII in a 70-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, lesion (arrow) remains hyperintense. (d) During equilibrium phase, lesion (arrow) has only slightly increased signal intensity compared with surrounding liver parenchyma. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is isointense compared with surrounding liver parenchyma. Lesion (arrow) is slightly hypointense during (f) portal venous phase and shows no change during (g) equilibrium phase.

 


View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (156K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (158K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2e. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (142K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2f. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (153K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2g. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segments IV and VIII in a 73-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is slightly hypointense compared with surrounding liver parenchyma, and adjacent small hyperintense area (arrowhead) consistent with intratumoral hemorrhage is seen. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, lesion (arrow) is hyperintense compared with adjacent liver parenchyma. (c) During portal venous phase, larger part of the lesion (arrow) is hypointense, consistent with rapid washout of contrast material. Only the outermost peripheral zone of the lesion remains hyperintense. (d) During equilibrium phase, central region of lesion (arrow) remains hypointense and peripheral region is isointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) is slightly hypointense. During (f) portal venous and (g) equilibrium phases, lesion (arrow) is still hypointense.

 


View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 


View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 


View larger version (154K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 


View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 


View larger version (147K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3e. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 


View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3f. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 


View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3g. Transverse dynamic fat-suppressed T1-weighted two-dimensional fast spoiled GRE MR images (150–170/1.4) of HCC in liver segment VII in a 69-year-old man with liver cirrhosis. (a) On precontrast image, lesion (arrow) is inhomogeneous: Central region is hypointense and periphery is slightly hyperintense compared with surrounding liver parenchyma. Note ascites (*) surrounding the liver. (b–d) Images obtained after administration of gadopentetate dimeglumine. (b) During hepatic arterial phase, central part of lesion shows rapid increase in signal intensity, and entire lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) During portal venous phase, central part of lesion (arrow) shows rapid washout of contrast material and is now hypointense, whereas peripheral region of lesion remains hyperintense. (d) During equilibrium phase, lesion (arrow) shows substantial loss in signal intensity and is now predominantly hypointense. (e–g) Corresponding images obtained after administration of ferucarbotran. (e) During hepatic arterial phase, lesion (arrow) has same characteristics as on the precontrast image: Central region is hypointense and periphery is slightly hyperintense. (f) During portal venous phase, lesion is isointense and, therefore, is hardly differentiable from surrounding liver parenchyma. (g) During equilibrium phase, enhancement pattern of lesion (arrow) remains the same as during portal venous phase.

 
On gadolinium-enhanced images obtained during the portal venous phase, only a few lesions remained hyperintense (10 of 36, 28%), whereas most lesions were hypointense (17 of 36, 47%) or isointense (nine of 36, 25%). On ferucarbotran-enhanced images obtained during the portal venous phase, many lesions were isointense (18 of 36, 50%) or hypointense (15 of 36, 42%), whereas a few lesions remained hyperintense (three of 36, 8%).

During the equilibrium phase, there was only one hyperintense lesion (3%) detected on gadolinium-enhanced images. The majority of the lesions were hypointense (21 of 36, 58%) and the remaining were isointense (14 of 36, 39%). On ferucarbotran-enhanced images obtained during the equilibrium phase, most lesions were isointense (22 of 36, 61%); the fraction of lesions that were hypointense decreased further (11 of 36, 31%), and a small fraction of hyperintense lesions remained stable (three of 36, 8%).

Enhancement Patterns
Ring enhancement surrounding the hepatic lesions was present more often after the administration of gadopentetate dimeglumine (seven lesions [19%] present during both the portal venous and equilibrium phases) than after the administration of ferucarbotran (one lesion [3%] present during the hepatic arterial phase, four lesions [11%] present during the portal venous phase, and one lesion [3%] present during the equilibrium phase). However, with use of ferucarbotran, ring enhancement was noted in at least one lesion at each examined phase.

The results of the McNemar test demonstrated that during the hepatic arterial and portal venous phases the distribution of enhancement patterns was very scattered. Therefore, there was no detectable trend in the signal intensity characteristics of lesions after administration of either contrast agent during these two imaging phases. To the contrary, statistically significant differences in signal intensities existed between gadolinium-enhanced and ferucarbotran-enhanced images obtained during the equilibrium phase: During the equilibrium phase, significantly more lesions were hypointense on gadolinium-enhanced images than on ferucarbotran-enhanced images, whereas significantly more lesions were isointense on ferucarbotran-enhanced images than on gadolinium-enhanced images (P < .001 for both).

Accordingly, the {kappa} statistic resulted in poor agreement between lesion signal intensities after administration of the two contrast agents during the hepatic arterial and portal venous phases ({kappa} = –0.02 and {kappa} = –0.06, respectively). During the equilibrium phase, there was only fair agreement between the lesion signal intensities after administration of the two contrast agents ({kappa} = 0.36).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
It is well known that contrast-enhanced MR imaging diagnostically outperforms unenhanced MR imaging in HCC depiction, especially in the depiction of small lesions. However, it is still debatable which contrast agent offers the highest accuracy in the diagnosis of HCC in the cirrhotic liver and in the noncirrhotic liver. The various studies concerning SPIO-enhanced and/or gadolinium-enhanced MR imaging for HCC have revealed controversial results. Sensitivity and specificity values reported in the literature for gadolinium-enhanced MR imaging in the detection of HCC have a very broad variance (12,20,21). This may be explained by differences in tumor size, differences with regard to the standard of reference for HCC, or differences in histologic grading, as well as by the fact that liver cirrhosis was not present as the underlying liver disease in all cases.

Nevertheless, there is general agreement that diagnostic performance of gadolinium-enhanced MR imaging with regard to the depiction of HCC in the cirrhotic liver is dependent on the tumor size (12,13,21). Sensitivity and specificity values are significantly higher for HCC lesions larger than 1.5 cm in diameter compared with those smaller than 1.5 cm in diameter (12). Similarly, there is also a broad variance in the diagnostic performance of SPIO with regard to the depiction of HCC in the cirrhotic and the noncirrhotic liver (12,20,22–26), with a similar tendency for better depiction of larger tumors than smaller tumors (12,13).

Another approach to increasing the diagnostic performance of MR imaging in the detection of HCC in the cirrhotic liver involves the combination of gadolinium- and SPIO-enhanced MR imaging (the so-called double-contrast MR imaging protocol), as proposed by Ward et al (13). Ward et al have shown that the sensitivity for detection of HCC may be substantially increased by using the double-contrast MR imaging protocol compared with using SPIO-enhanced imaging alone, especially for lesions less than 1 cm in diameter (13).

When performing diagnostic MR imaging for hepatic lesions, the diagnosis of HCC is based on a combination of characteristic imaging features visible on unenhanced as well as on contrast-enhanced images. On unenhanced T2-weighted fast spin-echo or single-shot fast spin-echo images, the majority of HCC lesions are hyperintense with regard to the surrounding liver parenchyma (27). However, a minority of HCC lesions may be isointense or slightly hypointense compared with the surrounding parenchyma (13,21,28). The signal intensity characteristics of HCC on unenhanced T1-weighted images are also variable, with the majority of the lesions being hypointense or isointense to the surrounding liver parenchyma (27). In some cases, HCC may be hyperintense on unenhanced T1-weighted MR images (24).

On gadolinium-enhanced T1-weighted MR images, early lesional signal enhancement during the hepatic arterial phase followed by a rapid total or partial washout of contrast material is considered a typical finding in HCC (29,30). Strong peritumoral ring enhancement is also described as a possible finding in HCC on gadolinium-enhanced MR images (29). Some subgroups of poorly differentiated HCCs may show slow and only slight increase in signal intensity and thus remain hypointense with regard to the surrounding liver parenchyma during the hepatic arterial phase (13,29).

When SPIO is used as a contrast agent for liver imaging, characteristic imaging features on postcontrast T2- or T2*-weighted images obtained during the accumulation phase are usually used for diagnosis of HCC. HCCs typically retain high signal intensity or show only a slight loss in signal intensity on SPIO-enhanced T2- or T2*-weighted images because SPIO shortens T2 or T2* relaxation times within liver tissue containing Kupffer cells but not within lesions that are lacking Kupffer cells. However, some well-differentiated HCCs may show substantial signal intensity loss on T2- or T2*-weighted images, as they may still contain a certain amount of Kupffer cells (31).

In most of the studies dealing with use of SPIOs for the imaging of HCC, only the T2 and T2* properties of SPIOs were used for diagnosis of HCC; this was mainly because of the pharmacologic properties of the product used. Until recently, ferumoxides (Feridex, Advanced Magnetics; Endorem, Guerbet) were the only marketed SPIOs in most countries. Although it is well known that ferumoxides have relatively high r1 values (up to four- or fivefold higher than those of gadolinium-based contrast agents) (14,16), these T1 effects could not be readily exploited for dynamic T1-weighted MR imaging. Bolus administration of ferumoxides has not been recommended because of the possible side effects (lumbar pain and cardiovascular problems, including dose-dependent hypotensive reactions) that positively correlate to the rate of injection; therefore, this compound has to be administered as a slow drip infusion (32,33). Nevertheless, the T1 effect of ferumoxides has been used for assessment of some liver lesions in delayed phases (15,34–37). Peritumoral ring enhancement was described as a frequent sign of malignant lesions, including HCC lesions, in the few studies performed by using T1-weighted MR imaging of hepatic lesions during the distributional phase from at least 10 minutes up to 1 hour after the intravenous slow-drip infusion of ferumoxides (34,35,37). In contrast with ferumoxides, ferucarbotran can be administered as a bolus, which enables dynamic T1-weighted MR imaging similar to dynamic imaging with gadolinium chelates.

Some authors (18,38–40) have investigated the imaging characteristics of hepatic lesions at dynamic T1-weighted MR imaging after the bolus injection of ferucarbotran. In the studies by these authors, small numbers of a broad variety of primary and secondary hepatic lesions, including HCC, were examined. However, an intraindividual comparison of the enhancement patterns with either ferucarbotran or gadolinium compounds was not performed in these studies.

In this study, we pursued an intraindividual comparison between the enhancement patterns of histologically proved HCC lesions at gadopentetate dimeglumine– and ferucarbotran-enhanced dynamic T1-weighted MR imaging. When gadopentetate dimeglumine was used, most of the HCC lesions (21 of 36, 58%) showed the classic enhancement pattern, with intense enhancement during the hepatic arterial phase followed by a rapid washout during the portal venous and equilibrium phases. Only a minority of the lesions were hypo- or isointense compared with the surrounding liver parenchyma during the hepatic arterial phase of dynamic T1-weighted gadopentetate dimeglumine–enhanced imaging. Contrarily, after administration of ferucarbotran, most of the lesions were isointense (18 of 36, 50%) or hypointense (11 of 36, 31%); only a few lesions were hyperintense during the hepatic arterial phase (seven of 36, 19%). During the equilibrium phase, the fraction of isointense lesions increased to more than 60%. These findings correspond to a more rapid washout of gadopentetate dimeglumine in comparison with ferucarbotran.

Peritumoral ring enhancement, which has been described as an important sign in distinguishing benign from malignant lesions at SPIO-enhanced MR imaging (18,29,34,35,37), was not a frequent finding in our study. After administration of gadopentetate dimeglumine, peritumoral ring enhancement was seen in approximately 20% of cases during the portal venous and equilibrium phases. After administration of ferucarbotran, only up to 11% of the lesions showed rim enhancement during the portal venous phase; however, there was at least one lesion with ring enhancement during each acquisition phase.

There are at least three possible explanations for the different enhancement patterns following administration of either gadopentetate dimeglumine or ferucarbotran. First, the different distribution volumes have to be considered. It is well known that within the first seconds after intravenous administration of gadopentetate dimeglumine, extracellular interstitial effects may overlap vascular effects (41,42), whereas SPIOs remain strictly intravascular for a much longer time after intravenous administration (43,44). Therefore, gadopentetate dimeglumine–enhanced image acquisition during the hepatic arterial phase might not strictly reflect arterial effects of the contrast agent within the hepatic lesion, whereas the enhancement patterns after administration of ferucarbotran might truly reflect vascular effects. In addition, the blood half-life of the contrast agents differs substantially. Gadopentetate dimeglumine has a blood half-life of 3–5 minutes at a dose of 0.1 mmol per kilogram of body weight (based on manufacturer data). The blood half-life of ferucarbotran is at least twice as long—approximately 10 minutes up to a dose of 40 µmol iron per kilogram of body weight (33). Also, there might be dose-related differences in enhancement characteristics, since ferucarbotran was used at a dose of approximately 0.01 mmol iron per kilogram of body weight, whereas gadopentetate dimeglumine is used at a 10-fold higher dose. However, increasing the dose of ferucarbotran might have converse effects, since administration of SPIO at doses higher than 10 µmol iron per kilogram can result in negative enhancement even on T1-weighted images (45).

The results of this study may have direct influence on the reading of liver images obtained by using ferucarbotran-enhanced dynamic T1-weighted imaging. As we have shown, the well-known enhancement patterns of HCC at gadolinium-enhanced dynamic T1-weighted MR imaging cannot be readily transferred to ferucarbotran-enhanced imaging, because the same HCC lesion may display different enhancement patterns when the vascular phases of these contrast agents are directly compared. Hence, in our daily clinical practice, the diagnosis of an HCC lesion is merely based on a lack of a signal intensity decrease on T2- or T2*-weighted images during the accumulation phase when ferucarbotran-enhanced MR imaging is used. Results of future studies should reveal whether the enhancement patterns of other primary liver lesions are also different on gadolinium- and SPIO-enhanced dynamic T1-weighted images.

We acknowledge the following limitations of our study: First, the number of investigated HCC lesions was relatively small. As discussed earlier, the relative doses of the contrast agents were different. However, we injected the maximal recommended dose of ferucarbotran in all patients. The results may have been affected by the fact that we used a fixed interval instead of bolus timing for acquisition of the dynamic T1-weighted images. We used this approach because of the small volume of ferucarbotran, which obviates the test bolus method for optimization of acquisition timing. Another limitation is the fact that we did not compare the diagnostic performance of each contrast agent in depicting HCC lesions. However, that was not the goal of this study; we aimed to compare the enhancement patterns of the two contrast agents by using T1-weighted GRE MR imaging at different vascular phases. Finally, we included mainly large HCC lesions (ie, >2 cm in diameter) and histologically well-differentiated HCC lesions, which may have resulted in an inclusion bias.

In conclusion, results of this study have shown that the enhancement patterns of HCC lesions on T1-weighted GRE MR images differ between gadolinium- and ferucarbotran-enhanced images. Hence, when ferucarbotran-enhanced MR imaging is performed for the assessment of HCC, lesion characterization should be based solely on imaging features of T2- and T2*-weighted images.


   ACKNOWLEDGMENTS
 
The authors thank D. W. Crook, MD, for his help in preparing the manuscript and B. Seifert, PhD, for his support in the statistical analysis.


   FOOTNOTES
 

Abbreviations: GRE = gradient-recalled echo • HCC = hepatocellular carcinoma • SPIO = superparamagnetic iron oxide

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, A.M.L., D.W.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, A.M.L., J.K.W., K.G.; clinical studies, J.K.W., K.G.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Badvie S. Hepatocellular carcinoma. Postgrad Med J 2000;76:4–11.[Abstract/Free Full Text]
  2. Ogunbiyi JO. Hepatocellular carcinoma in the developing world. Semin Oncol 2001;28:179–187.[CrossRef][Medline]
  3. Poon RT, Fan ST. Hepatectomy for hepatocellular carcinoma: patient selection and postoperative outcome. Liver Transpl 2004;10(suppl 2):S39–S45.
  4. Schwartz M. Liver transplantation in patients with hepatocellular carcinoma. Liver Transpl 2004;10(suppl 2):S81–S85.[CrossRef][Medline]
  5. Livraghi T, Goldberg SN, Lazzaroni S, Meloni F, Solbiati L, Gazelle GS. Small hepatocellular carcinoma: treatment with radio-frequency ablation versus ethanol injection. Radiology 1999;210:655–661.[Abstract/Free Full Text]
  6. Guglielmi A, Ruzzenente A, Battocchia A, Tonon A, Fracastoro G, Cordiano C. Radiofrequency ablation of hepatocellular carcinoma in cirrhotic patients. Hepatogastroenterology 2003;50:480–484.[Medline]
  7. Yamamoto J, Okada S, Shimada K, et al. Treatment strategy for small hepatocellular carcinoma: comparison of long-term results after percutaneous ethanol injection therapy and surgical resection. Hepatology 2001;34:707–713.[CrossRef][Medline]
  8. Llovet JM, Bruix J. Systematic review of randomized trials for unresectable hepatocellular carcinoma: chemoembolization improves survival. Hepatology 2003;37:429–442.[CrossRef][Medline]
  9. Poon RT, Fan ST, Lo CM, Liu CL, Wong J. Long-term survival and pattern of recurrence after resection of small hepatocellular carcinoma in patients with preserved liver function: implications for a strategy of salvage transplantation. Ann Surg 2002;235:373–382.[CrossRef][Medline]
  10. Scharf J, Hoffmann V, Lehnert T, Anselm H, Richter GM, Kauffmann GW. Pseudolesions at T1-weighted gradient-echo imaging after administration of superparamagnetic iron oxide: comparison with portal perfusion abnormalities at CT during arterial portography. Radiology 1998;207:67–72.[Abstract/Free Full Text]
  11. Denys A, Arrive L, Servois V, et al. Hepatic tumors: detection and characterization at 1-T MR imaging enhanced with AMI-25. Radiology 1994;193:665–669.[Abstract/Free Full Text]
  12. Pauleit D, Textor J, Bachmann R, et al. Hepatocellular carcinoma: detection with gadolinium- and ferumoxides-enhanced MR imaging of the liver. Radiology 2002;222:73–80.[Abstract/Free Full Text]
  13. Ward J, Guthrie JA, Scott DJ, et al. Hepatocellular carcinoma in the cirrhotic liver: double-contrast MR imaging for diagnosis. Radiology 2000;216:154–162.[Abstract/Free Full Text]
  14. Jung CW, Jacobs P. Physical and chemical properties of superparamagnetic iron oxide MR contrast agents: ferumoxides, ferumoxtran, ferumoxsil. Magn Reson Imaging 1995;13:661–674.[CrossRef][Medline]
  15. Kumano S, Murakami T, Kim T, et al. Using superparamagnetic iron oxide-enhanced MRI to differentiate metastatic hepatic tumors and nonsolid benign lesions. AJR Am J Roentgenol 2003;181:1335–1339.[Abstract/Free Full Text]
  16. Reimer P, Rummeny EJ, Daldrup HE, et al. Clinical results with Resovist: a phase 2 clinical trial. Radiology 1995;195:489–496.[Abstract/Free Full Text]
  17. Oudkerk M, van den Heuvel AG, Wielopolski PA, Schmitz PI, Borel Rinkes IH, Wiggers T. Hepatic lesions: detection with ferumoxide-enhanced T1-weighted MR imaging. Radiology 1997;203:449–456.[Abstract/Free Full Text]
  18. Reimer P, Muller M, Marx C, et al. T1 effects of a bolus-injectable superparamagnetic iron oxide, SH U 555 A: dependence on field strength and plasma concentration—preliminary clinical experience with dynamic T1-weighted MR imaging. Radiology 1998;209:831–836.[Abstract/Free Full Text]
  19. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174.[CrossRef][Medline]
  20. Matsuo M, Kanematsu M, Itoh K, et al. Detection of malignant hepatic tumors: comparison of gadolinium- and ferumoxide-enhanced MR imaging. AJR Am J Roentgenol 2001;177:637–643.