| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gastrointestinal Imaging |
2-cm) Early-Enhancing Hepatic Lesions with Serial Contrast-enhanced Dynamic MR Imaging1
1 From the Department of Radiology, Yamaguchi University School of Medicine, 1-1-1 Minamikogushi, Ube, Yamaguchi 755-8505 Japan. Received December 3, 2001; revision requested February 14, 2002; revision received April 5; accepted June 24. Address correspondence to A.S. (e-mail: ayamechan51@hotmail.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
2 cm) early-enhancing hepatic lesions in cirrhotic liver disease with serial multiphasic contrast material–enhanced dynamic magnetic resonance (MR) imaging.
MATERIALS AND METHODS: The study population included 208 patients with cirrhosis (n = 162) or chronic hepatitis (n = 46) who underwent dynamic contrast-enhanced MR imaging at least twice during the clinical course, with at least a 12-month interval between the initial and latest MR examinations. Initial images were evaluated for the presence and shape of small (2 cm) early-enhancing hepatic lesions at arterial-phase contrast-enhanced MR imaging. If a small early- enhancing lesion was seen on the initial image, the latest follow-up image was reviewed to assess the serial changes in the features of a specific lesion.
RESULTS: One hundred fifty-eight small early-enhancing lesions were detected in 75 (36%) of the 208 patients on the initial MR images. Of those 158 lesions, 104 were round or oval, 30 wedge-shaped, 18 geographic (irregularly shaped), and six triangular. Among the 104 round or oval lesions, 54 (52%) disappeared or decreased in size on the latest MR images and were considered definite pseudolesions. Twenty-nine (28%) of the 104 round or oval lesions were classified as hepatocellular carcinoma (HCC) on the basis of their interval growth or pathologic confirmation. The remaining 21 (20%) lesions were stable in size and appearance and considered probable pseudolesions. For the other three shapes, 73% of wedge-shaped, 78% of geographic, and 67% of triangular lesions disappeared or decreased in size.
CONCLUSION: Small early-enhancing hepatic lesions in patients with cirrhosis usually showed no interval growth or disappeared during serial contrast-enhanced dynamic MR imaging. Even though these lesions are round or oval, they may more frequently be pseudolesions than HCCs.
© RSNA, 2003
Index terms: Hepatitis, 761.291 • Liver, cirrhosis, 761.794 • Liver, MR, 761.121411, 761.121412 • Liver neoplasms, MR, 761.323 • Magnetic resonance (MR), contrast media • Shunts, arterioportal
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
2 cm) early enhancing hepatic lesions seen at serial multiphasic contrast-enhanced dynamic MR imaging in patients with chronic liver disease. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Consequently, this study population included 208 patients with cirrhosis (n = 162) or chronic hepatitis (n = 46). There were 154 men and 54 women aged 26–83 years (mean, 61 years). All patients were followed up with MR imaging for at least 12 months (range, 12–64 months; mean, 32 months). Chronic hepatitis or cirrhosis was related to viral infection (hepatitis B [n = 28], hepatitis C [n = 136], or both [n = 4]), alcohol abuse (n = 10), or autoimmune hepatitis (n = 2). In 28 patients, the cause of chronic hepatitis or cirrhosis was unknown (n = 10) or could not be determined owing to insufficient clinical data (n = 18). The diagnosis of chronic hepatitis or cirrhosis was made by means of percutaneous liver biopsy in 14 patients and on the basis of the clinical course and evaluation (including liver function tests) in 194. One hundred fifty-eight patients underwent MR imaging more than three times during the follow-up period; the images obtained at the initial and latest examination were used for image analysis.
MR Imaging Technique
MR imaging was performed with a 1.5-T unit (Magnetom H-15 or Vision; Siemens Medical Systems, Erlangen, Germany) by using a whole-body coil or phased-array torso coil. Before MR imaging, radiologists explained the purpose of the examination to each patient and obtained informed consent. All patients underwent transverse T1- and T2-weighted MR imaging. T1-weighted imaging included at least one of the following sequences: conventional spin echo (SE) (400–600/12–15 [repetition time msec/echo time msec]), in-phase gradient echo (GRE) (70–210/4.0–4.8 or 9.0, 70°–90° flip angle), and opposed-phase GRE (80–210/2.3–2.7 or 6.0–7.0, 60°–90° flip angle). T2-weighted imaging included conventional SE (2,000–3,000/90–100), non–breath-hold fast SE (3,000–4,000/96–102 [effective echo time]) with or without fat suppression, or breath-hold fast SE (2,500–3,000/70 or 93 [effective echo time]). Multiphasic contrast-enhanced dynamic MR imaging was performed before and after the administration of gadopentetate dimeglumine (Magnevist; Japan-Schering, Osaka, Japan) by using a two-dimensional or three-dimensional T1-weighted GRE sequence (two-dimensional: 120–210/2.3–6.0, 60°–90° flip angle; three-dimensional: 6.0/2.4, 20° flip angle) with or without fat suppression. Superior and inferior presaturation pulses were applied in two-dimensional dynamic MR imaging to reduce the occurrence of flow artifacts.
After the acquisition of unenhanced GRE images, 0.1 mmol per kilogram of body weight of contrast material was injected intravenously as a rapid bolus while the patient was in the bore of the magnet. This was followed by a 10-mL normal saline flush with (n = 37) or without (n = 379) the use of a power injector (Spectris; Medrad, Indianola, Pa). The decision to administer contrast material intravenously was based on physician judgment at the examination. Contrast-enhanced dynamic imaging was initiated at 25 seconds, 60 seconds, and 3 minutes after the start of the bolus injection to obtain multiphasic images (arterial, portal, and equilibrium phases). Finally, delayed-phase contrast-enhanced GRE images were obtained with fat suppression at 5–15 minutes after the injection of contrast material. The imaging matrix was 256 x 128, 160, or 192 pixels for SE and GRE sequences, 256 x 256 pixels for non–breath-hold fast SE sequences, and 256 x 192 for breath-hold fast SE sequences. We usually used a rectangular field of view to reduce the number of phase-encoding views. We used 7–12-mm-thick sections and a section gap of 2 mm or less.
Image Interpretation
All images were reviewed retrospectively and independently by two radiologists (A.S. and K.I.) who had more than 10 years of experience in liver MR imaging. They did not know the final diagnosis. MR images were evaluated for the presence of small (2 cm) early enhancing hepatic lesions on the arterial-phase contrast-enhanced dynamic MR images. The two readers disagreed regarding the presence of 18 early enhancing lesions. These disagreements were resolved by means of consensus after discussion. Seven of the 18 lesions were included in this study. Eleven of the 18 lesions were excluded because they were judged by consensus reading to represent small vessels or a component of heterogeneous parenchymal enhancement throughout the liver.
The review procedure was performed at two separate sessions. At the first session, the initial MR images were reviewed for the presence of small early enhancing hepatic lesions. The size and shape of early enhancing lesions were recorded. The shape of the lesion was categorized as round or oval, wedge shaped, triangular, or geographic (irregularly shaped). The signal intensity on T2-weighted and delayed-phase contrast-enhanced images was also recorded when visible. In the second session, the latest MR images were compared with the initial MR images to assess the serial changes in the features of specific lesions. When the early enhancing lesion was enlarged at any time or pathologically proved to be HCC during the follow-up period, it was considered HCC. When the early enhancing lesion disappeared or decreased in size at follow-up imaging, it was considered a definite pseudolesion. When the early enhancing lesion was stable in size and appearance for at least 12 months, it was considered a probable pseudolesion.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
2.0 cm) early enhancing lesions were detected in 75 (36%) of the 208 patients during the arterial phase of contrast-enhanced dynamic MR imaging. Among these 158 lesions, 22 were slightly hyperintense to the surrounding liver parenchyma on unenhanced T1-weighted MR images; however, early enhancement was visually obvious in all lesions. The mean lesion size was 0.96 cm (range, 0.2–2.0 cm). One hundred four (66%) lesions were round or oval, 30 (19%) were wedge shaped, 18 (11%) were geographic, and six (4%) were triangular. Among the 104 round or oval lesions, 54 (52%) disappeared (n = 48) or decreased in size (n = 6) on the latest MR images; these were considered definite pseudolesions (Figs 1–3). Twenty-nine (28%) round or oval lesions were classified as HCCs (Fig 4) on the basis of their interval growth (n = 18) or pathologic confirmation (n = 11). The remaining 21 (20%) lesions were stable in size and appearance on the latest MR images (Fig 5) and were considered probable pseudolesions.
|
|
|
|
|
|
|
|
|
|
Of the 30 wedge-shaped lesions, 22 (73%) decreased in size or disappeared and were classified as definite pseudolesions. Eight (27%) were stable at follow-up and considered to be probable pseudolesions. Of the 18 geographic lesions, 14 (78%) disappeared or decreased in size and one (5%) showed interval growth. The remaining three (17%) lesions did not show any change in size or appearance. Of the six triangular lesions, four (67%) disappeared during the follow-up period and one (17%) increased in size; no interval change was observed in the remaining lesion. The overall frequency of small early enhancing hepatic lesions that disappeared or decreased in size or were stable during follow-up and were classified as pseudolesions was 80% (127 of 158 lesions).
On T2-weighted images, 14 (45%) of the 31 small hypervascular HCCs had slightly high signal intensity; the remaining 17 (55%) HCCs were not visible. Among 127 early enhancing pseudolesions, 123 (97%) were not depicted on T2-weighted images. Sixteen (52%) of the 31 small hypervascular HCCs were detected on delayed-phase contrast-enhanced images; the remaining 15 (48%) were not visible. One hundred eighteen (93%) of the 127 early enhancing pseudolesions were not visible and nine (7%) showed prolonged enhancement on delayed-phase contrast-enhanced images.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Multiphasic dynamic MR imaging of the whole liver after the injection of a bolus of gadolinium chelate has been reported as a highly sensitive method for detecting hypervascular HCCs in chronic liver damage because they are shown as early enhancing lesions on the arterial-phase images (2,12,13). In most previous studies, however, early enhancing pseudolesions as false-positive lesions have not been evaluated, resulting in the undetermined specificity of the lesions. We believe it is important to recognize that small early enhancing round or oval hepatic lesions in patients with cirrhosis or chronic hepatitis can often be pseudolesions to avoid unnecessary liver biopsy or treatment.
The small early enhancing pseudolesions seen on the arterial-phase dynamic MR images in this study may have been caused by localized hemodynamic change in blood flow in the hepatic parenchyma. Occlusion of small hepatic veins can occur in cirrhotic livers, resulting in increased sinusoidal pressure and reversing of the pressure gradient between the sinusoidal and portal veins. As a result, the affected area is predominantly perfused by arterial blood flow with retrograde filling of the small portal veins as the draining veins (14,15). In an experimental study of rat cirrhosis, hypertrophy of the peribiliary plexus and direct anastomosis in the peripheral area of the liver were shown at electron microscopy to be major routes of arterioportal communication (16).
In addition, temporarily decreased blood flow of the small portal branches in cirrhosis can cause a compensatory increase in the hepatic arterial flow within the affected liver parenchyma and a functional arterioportal shunt. The wedge-shaped or triangular pseudolesions in this study were presumably produced by small arterioportal shunts in the liver, as described in previous reports (8–10). The focally increased arterial inflow and decreased portal perfusion caused by the small arterioportal shunt may induce early enhancing pseudolesions. Round or oval early enhancing pseudolesions also may be explained by small focal arterioportal shunts, depending on their three-dimensional location in the plane of transverse MR images. These arterioportal shunts may appear to be wedge shaped when they reside squarely in the plane of section. In addition, round or oval early enhancing pseudolesions can occur when the portal or hepatic veins that do not reach the hepatic capsule (17).
The observation of serial changes in the MR imaging features of early enhancing lesions (disappearance of the lesion or a decrease in lesion size) during follow-up will support our speculation that early enhancing pseudolesions may be a result of an hepatic perfusional anomaly. Baron et al (18) performed biphasic contrast-enhanced dynamic computed tomography (CT) of the cirrhotic liver and found less attenuated but scattered diffuse foci of irregular enhancement similar to that seen in our small early enhancing round pseudolesions. Baron et al found no pathologic correlate at visual inspection of the transplant specimen to account for these diffuse foci.
The authors of several studies have reported the presence of nodular early enhancing pseudolesions on arterial-phase contrast-enhanced dynamic MR images; the lesions were diagnosed with arteriography as being small arterioportal shunts (8–10). Meticulous correlation of our findings with those of hepatic arteriography, CT during arterial portography, and CT arteriography would have been interesting in our patient population.
In this study, delayed-phase contrast-enhanced MR images and T2-weighted images showed no focal areas with abnormal signal intensity that corresponded to the location with early enhancement in 93%–97% of the early enhancing pseudolesions. This nonvisibility may be attributed to the nature of early enhancing pseudolesions, which are potentially caused by hepatic perfusion abnormalities such as small arterioportal shunts without pathologic alteration. In 31 small hypervascular HCCs, 48% of the lesions were detected only on arterial-phase images as an area of transient focal enhancement without any distinguishable abnormalities on T2-weighted and delayed-phase contrast-enhanced MR images. This finding indicates that there was considerable overlap between the MR imaging features of hypervascular HCCs and those of early enhancing pseudolesions, resulting in the difficulty in definitive differentiation between them.
It was reported that single-level dynamic CT during hepatic arteriography has the potential to enable the differentiation of hypervascular HCCs from hypervascular pseudolesions on the basis of the presence of coronal stain of adjacent liver (11). However, this technique is invasive, expensive, limited to the assessment of a solitary lesion, and not suitable for the routine clinical examination of the whole liver. As an alternative technique, development of high-speed CT or MR imaging to image the whole liver repeatedly during short time intervals (2–4 seconds) would be necessary in the future.
This study has some limitations. It is a retrospective study and, consequently, there is a variability to MR imaging parameters that were used because imaging protocols changed during the many years of the study. This variation may have influenced the visualization of some of the small lesions. However, imaging protocols have been changed according to technical improvements, and standardized pulse sequences optimized for liver imaging have been selected. For instance, for two-dimensional dynamic MR imaging, we used 7–12-mm-thick sections with a 2–0-mm gap, which may have resulted in nonvisualization of the small lesions. Three-dimensional dynamic MR imaging, which was used in the later protocols, would have enabled better detection.
The issue of arterial imaging phase should also be discussed. In this study, arterial-phase imaging was performed after a fixed imaging delay after the start of intravenous contrast material injection and was possibly influenced by the differences in circulatory physiology among individuals (19,20). In addition, a power injector was not used in most patients, and the decision to administer contrast material intravenously (eg, left-sided injection) (21) was based on physician judgment. For meticulous analysis, it would be important to customize the timing of arterial-phase acquisitions for each individual. Otherwise, multiple successive early image acquisitions of the whole liver are helpful for obtaining optimal arterial-phase images (22).
In the current study, we did not focus on the diagnostic value of MR imaging in the characterization of small early enhancing hepatic lesions detected during the arterial phase of dynamic imaging. However, this study was not designed to assess the ability of MR imaging to enable differentiation of small hypervascular HCCs from early enhancing hepatic pseudolesions. Instead, the purpose was to assess the frequency and clinical importance of small early enhancing hepatic lesions in patients with cirrhosis or chronic hepatitis detected in clinical MR practice.
Another limitation of this study is that probable pseudolesions were determined on the basis of their stability at follow-up MR imaging, not by means of histologic confirmation. Pathologic proof of small early enhancing pseudolesions presumably due to an arterioportal shunt, however, would not be practical in the clinical setting. On the basis of our criterion of stability for at least 12 months and the relatively long follow-up available (mean, 23 months), we believe that it is unlikely that hypervascular HCCs were incorrectly categorized as probable pseudolesions, although the possibility of an extremely slow-growing HCC cannot be excluded completely.
In conclusion, small early enhancing hepatic lesions in patients with cirrhosis or chronic hepatitis are often seen on arterial-phase contrast-enhanced dynamic MR images. These small early enhancing lesions may more frequently be pseudolesions than HCCs even though they are round or oval, because these lesions usually disappeared, decreased in size, or showed no interval growth on the serial contrast-enhanced dynamic MR images obtained during long-term clinical follow-up.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, A.S., K.I.; study concepts, A.S.; study design, K.I.; literature research, K.S., T.F.; clinical studies, A.S.; data acquisition, S.K., K.S.; data analysis/interpretation, A.S.; statistical analysis, S.K.; manuscript preparation, A.S.; manuscript definition of intellectual content, K.I.; manuscript editing, A.S., K.I.; manuscript revision/review, K.I.; manuscript final version approval, A.S., N.M.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. M. Willatt, H. K. Hussain, S. Adusumilli, and J. A. Marrero MR Imaging of Hepatocellular Carcinoma in the Cirrhotic Liver: Challenges and Controversies Radiology, May 1, 2008; 247(2): 311 - 330. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Ronzoni, D. Artioli, R. Scardina, L. Battistig, E. Minola, S. Sironi, and A. Vanzulli Role of MDCT in the Diagnosis of Hepatocellular Carcinoma in Patients with Cirrhosis Undergoing Orthotopic Liver Transplantation Am. J. Roentgenol., October 1, 2007; 189(4): 792 - 798. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Choi, D. G. Mitchell, S. K. Verma, D. Bergin, V. J. Navarro, A. B. Malliah, C. McGowan, H.-W. L. Hann, and S. K. Herrine Hepatocellular Carcinoma with Indeterminate or False-Negative Findings at Initial MR Imaging: Effect on Eligibility for Curative Treatment Initial Observations Radiology, September 1, 2007; 244(3): 776 - 783. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kanematsu, S. Goshima, H. Kondo, Y. Tsuge, R. Yokoyama, K. Kajita, M. Onozuka, Y. Suzuki, M. Van Cauteren, and N. Moriyama Gadolinium-Enhanced Multiphasic 3D MRI of the Liver with Prospective Adaptive Navigator Correction: Phantom Study and Preliminary Clinical Evaluation Am. J. Roentgenol., April 1, 2007; 188(4): W309 - W316. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kanematsu, S. Goshima, H. K. Hussain, T. D. Johnson, H. V. Nghiem, and I. R. Francis Does T2-weighted MR Imaging Really Add No Value in Detection and Characterization of Focal Lesions in Cirrhotic Liver? * Dr Hussain and colleagues respond: Radiology, February 1, 2005; 234(2): 638 - 640. [Full Text]
|
|
|
|
|
|
K. Ito, T. Fujita, A. Shimizu, S. Koike, K. Sasaki, N. Matsunaga, S. Hibino, and M. Yuhara Multiarterial Phase Dynamic MRI of Small Early Enhancing Hepatic Lesions in Cirrhosis or Chronic Hepatitis: Differentiating Between Hypervascular Hepatocellular Carcinomas and Pseudolesions Am. J. Roentgenol., September 1, 2004; 183(3): 699 - 705. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
