HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) CME Test (opens in a new window) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hanna, R. F. Articles by Sirlin, C. B. Search for Related Content PubMed PubMed Citation Articles by Hanna, R. F. Articles by Sirlin, C. B. Related Collections Magnetic Resonance Imaging Gastrointestinal Radiology

Cirrhosis-associated Hepatocellular Nodules: Correlation of Histopathologic and MR Imaging Features¹

¹ From the Department of Radiology, Division of Body Imaging (R.F.H., D.A.A., N.K., C.B.S.), and Department of Pathology (S.C.E., M.R.P.), University of California, San Diego, UCSD Liver Center, 200 W Arbor Dr, San Diego, CA 92103-8756. Recipient of a Certificate of Merit award for an education exhibit at the 2004 RSNA Annual Meeting. Received May 2, 2005; revision requested September 13; final revision received August 7, 2007; accepted August 21. The authors discuss an investigational or unlabeled use of a commercial product, device, or pharmaceutical that has not been approved for such purpose by the FDA. C.B.S. has received research grants from Bayer and General Electric; all remaining authors have no financial relationships to disclose. Address correspondence to C.B.S. (e-mail: csirlin{at}ucsd.edu).

	Abstract

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
Cirrhotic livers are characterized by advanced fibrosis and the formation of hepatocellular nodules, which are classified histologically as either (a) regenerative lesions (eg, regenerative nodules, lobar or segmental hyperplasia, focal nodular hyperplasia) or (b) dysplastic or neoplastic lesions (eg, dysplastic foci and nodules, hepatocellular carcinomas). The differentiation of these lesions is important because regenerative nodules are benign, whereas dysplastic and neoplastic nodules are premalignant and malignant, respectively. However, their accurate characterization may be difficult even at histopathologic analysis. Differential diagnosis may be facilitated by comparing the clinical and pathologic findings with radiologic imaging features; in particular, nodule size, vascularity, hepatocellular function, and Kupffer cell density assessed at magnetic resonance (MR) imaging are suggestive of the correct diagnosis. MR imaging is more useful than computed tomography for such assessments because it provides better soft-tissue contrast and a more nuanced depiction of different tissue properties. Moreover, a wider variety of contrast agents is available for use in MR imaging. Familiarity with the MR imaging characteristics of cirrhosis-associated hepatocellular nodules is therefore important for optimal diagnosis and management of cirrhotic disease.

© RSNA, 2008

	LEARNING OBJECTIVES

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
After reading this article and taking the test, the reader will be able to:

• Recognize cirrhotic nodules on unenhanced and contrast-enhanced MR images.
  
• Correlate gross pathologic and histologic findings with MR imaging features of cirrhotic nodules.
  
• Differentiate benign cirrhotic nodules from premalignant and malignant ones.
  

	Introduction

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
Liver cirrhosis is characterized by irreversible remodeling of the hepatic architecture with bridging fibrosis and a spectrum of hepatocellular nodules (1). Cirrhosis-associated hepatocellular nodules result from the localized proliferation of hepatocytes and their supporting stroma in response to liver injury (2). Most are benign regenerative nodules; however, regenerative nodules may progress along a well-described carcinogenic pathway to become dysplastic nodules or hepatocellular carcinomas (3). The imaging evaluation of cirrhosis-associated hepatocellular nodules therefore is important for their optimal management.

The article reviews the current classification of cirrhosis-associated hepatocellular nodules and surveys the magnetic resonance (MR) imaging techniques that may be used to evaluate the cirrhotic liver. The MR imaging characteristics (tissue-specific signal intensities and contrast agent–induced enhancement patterns) of various kinds of nodules are correlated with the gross pathologic and histologic findings.

	Nodule Classification

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
In 1995, an International Working Party panel published a set of guidelines for classifying and describing nodular hepatocellular lesions (2). According to those guidelines, there are two categories of hepatocellular nodules: (a) regenerative lesions and (b) dysplastic or neoplastic lesions (Table 1).

Dysplastic and neoplastic cirrhotic lesions include dysplastic foci and nodules, hepatocellular carcinomas, and some hepatocellular adenomas, although the latter typically occur in the noncirrhotic liver.

Thus, there are four classes of lesions that characteristically are found in the cirrhotic liver: regenerative nodules, dysplastic foci, dysplastic nodules, and hepatocellular carcinomas. These lesions are referred to collectively as cirrhosis-associated hepatocellular nodules.

Regenerative Nodules
Regenerative nodules form in response to necrosis, altered circulation, or other stimuli (2). They may be monoacinar or multiacinar, depending on whether they contain one or more terminal portal tracts. Regenerative nodules also may be classified according to size as either micronodules (<3 mm) or macronodules (3 mm). Giant regenerative nodules with a diameter of 5 cm have been described, but they are rare (2).

Dysplastic and Neoplastic Nodules
Cirrhosis-associated dysplastic and neoplastic nodules are composed of hepatocytes that display histologic characteristics of abnormal growth caused by a presumed or proved genetic alteration (2). Lesions with dysplastic features that do not satisfy the histologic criteria for malignancy or invasion are described as either (a) dysplastic foci (<1 mm in diameter) or (b) dysplastic nodules (1 mm in diameter). Dysplastic nodules usually occur in the setting of cirrhosis and may be classified as low or high grade, according to the degree of dysplasia.

The diagnostic differentiation of dysplastic nodules from other cirrhosis-associated hepatocellular nodules may be difficult even at histopathologic analysis, and the use of molecular genetics–based techniques may be necessary (2,4). Moreover, the clinical relevance of dysplastic nodules is unclear. Although these nodules may transform over time into hepatocellular carcinomas, the rate of that transformation is relatively slow. According to the latest guidelines from the American Association for the Study of Liver Diseases, dysplastic nodules should not be treated or managed as cancers, and patients with known or suspected dysplastic nodules should not be monitored more aggressively than patients without such nodules (5). Because of the challenges of diagnosing dysplastic nodules on the basis of histologic analysis and the uncertain clinical relevance of these lesions, the United Network of Organ Sharing does not require the reporting of dysplastic nodules in prospective liver transplants or explanted livers (6). Nevertheless, an understanding of the histologic and imaging features of dysplastic nodules is important because these lesions may cause diagnostic confusion at MR imaging, with resultant errors in interpretation and management.

Hepatocellular Carcinomas.— Hepatocellular carcinomas are malignant neoplasms composed of dedifferentiated hepatocytes. The lesions generally are described either as small (<2 cm in diameter) or large (2 cm in diameter). The classic system of macroscopic classification of hepatocellular carcinomas, in use since 1901, includes three major types: nodular, massive, and diffuse (7). Nodular hepatocellular carcinomas are small lesions with distinct margins. Massive carcinomas are larger than nodular ones and may consist of several confluent small lesions, a single dominant lesion, or a combination thereof. Diffuse hepatocellular carcinoma is characterized by multiple infiltrating tumors (7). In 1987, Kanai and colleagues described three major subtypes of nodular hepatocellular carcinomas: a single nodule, a single nodule with extranodular growth, and multiple contiguous nodules (8,9). The tumors are scored by using the Edmondson and the World Health Organization grading systems (7,9).

Siderotic Nodules.— Although the term siderotic nodule is not included in the International Working Party lexicon, it is mentioned here because it appears commonly in the radiology literature. The term was coined by radiologists to describe cirrhosis-associated nodules with high levels of endogenous iron. The longer term siderotic regenerative nodule was once in vogue, but at histologic analysis siderotic nodules may be regenerative or dysplastic, and no imaging finding can help differentiate between the two types; therefore, the term siderotic nodule is now favored (10). To our knowledge, siderotic nodules rarely, if ever, are malignant (11). This observation lends support to the hypothesis that dysplastic siderotic nodules lose their ability to accumulate endogenous iron as they undergo malignant transformation.

Pathway of Carcinogenesis.— Cirrhosis is the most common predisposing factor for hepatocellular carcinoma, and an estimated 80%–90% of patients with hepatocellular carcinoma in the United States have cirrhosis (12). The risk factors for hepatocellular carcinoma are summarized in Table 2.

	Nodule Characterization

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
Size
As lesions grow, the likelihood of high-grade dysplasia or frank malignancy increases. As a general rule, lesions with a diameter of less than 2 cm are more likely to be benign than malignant and, if malignant, are usually well differentiated (16,17). By comparison, lesions with a diameter of more than 2 cm are more likely to be malignant than benign and tend to be characterized by moderate to poor differentiation.

Vascularity
Regenerative nodules and low-grade dysplastic nodules are predominantly portally perfused and, after gadolinium administration, show enhancement similar to that of the surrounding liver (18,19). As dedifferentiation progresses within these nodules, angiogenic pathways are activated that induce new vessel formation, which manifests as an increased density of unpaired arteries and capillary units (16,18,20). This development leads to an increasing shift from predominantly venous perfusion to predominantly arterial perfusion as low-grade dysplastic nodules and hepatocellular carcinomas become high-grade lesions (18). The increasingly dedifferentiated nodules appear more markedly enhanced on early arterial phase images obtained after the intravenous injection of a contrast agent, with more pronounced washout on venous phase images and equilibrium phase images (21,22). The major shift in angiogenesis typically occurs during the transition from low-grade to high-grade dysplasia (18).

Hepatocellular Function
Regenerative nodules generally have normal hepatocellular function and therefore demonstrate avid uptake of hepatocellular contrast agents. As dedifferentiation proceeds, the number of expressed organic ion transporters decreases, with a resultant progressive reduction in the uptake of hepatocellular agents (22–25).

Kupffer Cell Density
The density of Kupffer cells within regenerative lesions is similar to that in the surrounding nonneoplastic hepatic parenchyma. The cell density is visible at contrast-enhanced imaging because Kupffer cells avidly accumulate particulate agents through phagocytic mechanisms. With dedifferentiation, the Kupffer cell density changes; however, such changes are unpredictable during the early stages of carcinogenesis. According to empirically derived values reported in the literature, dysplastic nodules and well-differentiated hepatocellular carcinomas have variable Kupffer cell densities, ranging from diminished to elevated levels (26–28). Moderately and poorly differentiated hepatocellular carcinomas tend to have a diminished Kupffer cell density (26,27).

Importance of Clinical History and Laboratory Testing
The clinical history and laboratory test results provide crucial information for the assessment and characterization of cirrhosis-associated hepatocellular nodules (Table 4) (29–31). The diagnostic value of imaging is greatest when the images are interpreted with consideration given to the available clinical and laboratory data.

In comparison with computed tomography, MR imaging generates better soft-tissue contrast, is useful for assessing a larger number of tissue properties, and provides greater sensitivity to contrast media. In addition, a wider variety of contrast agents are available for use in MR imaging. As a result, MR imaging has emerged as an important modality for the assessment of cirrhosis-associated hepatocellular nodules (32).

Unenhanced Acquisitions.— In general, the pulse sequences most commonly used at unenhanced MR imaging for lesion characterization are those that produce T1, T2, and T2* weighting. Unenhanced imaging may be performed with multiple pulse sequences and variable parameters. Chemical fat-saturation sequences and gradient-recalled echo (GRE) sequences with out-of-phase and in-phase image acquisitions are helpful for assessing hepatic or intralesional steatosis. In addition, diffusion-weighted sequences are used at some institutions to evaluate hepatocellular nodules (33).

Contrast-enhanced Acquisitions.— Three classes of MR contrast agents are available for assessing cirrhosis-associated hepatocellular nodules: gadolinium chelates with low molecular weight, hepatocellular agents, and superparamagnetic iron oxide (SPIO) particles (34–38).

Low-molecular-weight gadolinium chelates are extracellular paramagnetic contrast agents that generate T1 shortening and provide information about tissue vascularity. To assess vascularity, T1-weighted images are acquired before and in multiple phases after the administration of the gadolinium chelate. Volumetric three-dimensional (3D) fat-saturated spoiled GRE sequences are well suited for this purpose. Critical dynamic imaging phases include the hepatic arterial phase, in which the acquisition of the center of k-space coincides with the peak arterial perfusion of hepatic nodules. Portal venous phase images are generally acquired 20–30 seconds after gadolinium administration. Recently, several investigators have emphasized the importance of acquiring delayed venous phase and equilibrium phase images 180–240 seconds after contrast agent injection, to better assess venous washout (22,24,39).

However, the disadvantages of using gadolinium chelates must be borne in mind. Precise timing and patient cooperation during image acquisition are crucial because gadolinium-related enhancement is transient. Even if the acquisition of images is precisely timed, well-differentiated hepatocellular carcinomas and some dysplastic nodules may be portally perfused and hypoenhanced or isoenhanced during the arterial phase, thus evading detection. Benign cirrhotic tissue may be hyperenhanced because of altered hemodynamics and may obscure underlying hepatocellular carcinomas. In addition, benign arteriovenous shunts and active inflammation may cause early enhancement mimicking that in hepatocellular carcinoma (5).

Mixed extracellular and hepatocellular agents constitute a specific subcategory of gadolinium chelates. These agents are useful not only for evaluating lesion vascularity, like other gadolinium chelates, but also for assessing hepatocellular function on images acquired with an appropriate delay. Two such agents have been manufactured commercially: gadobenate dimeglumine (Gd-BOPTA) and gadoxetic acid disodium (Gd-EOB-DTPA). Gadobenate dimeglumine is the only mixed extracellular and hepatocellular agent currently approved for use in the United States.

Up to 5% of the administered dose of gadobenate dimeglumine is selectively taken up by functioning hepatocytes and excreted into the bile by the multispecific organic ion transporter shared by bilirubin. Although initial biliary excretion of the agent is noticeable on delayed T1-weighted images at 5 minutes after injection, reliable characterization of hepatocellular function in the cirrhotic liver usually requires delays of 30–120 minutes (22,24). Thus, although a single injection of gadobenate dimeglumine permits assessment of lesion vascularity and hepatocellular function, two imaging sessions are typically required if both types of data are desired.

SPIO particles, which have a volume-weighted mean diameter of 25–250 nm, are phagocytosed by Kupffer cells in the liver and spleen. At MR imaging, SPIO particles cause shortening of T2* and, to a lesser degree, T2. In the cirrhotic liver, they accumulate in tissues that contain Kupffer cells (eg, in regenerative nodules, some dysplastic nodules, and regions of surrounding liver parenchyma), causing signal loss in those regions on T2-weighted images and T2*-weighted images. Most hepatocellular carcinomas lack Kupffer cells, do not accumulate SPIO particles, and therefore appear hyperintense relative to the liver parenchyma on T2-weighted images and T2*-weighted images obtained after the administration of SPIO.

Two approaches are available to assess phagocytic function by using SPIO. The first is to acquire images both before and after the contrast agent infusion. This method permits a direct comparison of unenhanced and SPIO-enhanced images and an unequivocal assessment of the degree of accumulation of the contrast agent in the liver and nodules, and it has been shown to reduce false-positive findings caused by blood vessels (40). Another advantage is that the unenhanced images may show evidence of excessive endogenous iron, which is a contraindication against subsequent SPIO administration. However, because the administration of SPIO particles is time consuming and because dynamic imaging requires two separate acquisition sessions, this approach may be impractical in the clinical setting. An alternative is to administer the SPIO particles before the initial image acquisition (ie, to acquire only SPIO-enhanced images). This approach is based on the assumption that lesions with high signal intensity on T2- or T2*-weighted images have a reduced Kupffer cell density. Although this method is faster and easier than dynamic imaging, it increases the risk that SPIO particles may be administered to patients in whom there is an excess of endogenous iron. Moreover, it does not allow the direct assessment of SPIO particle accumulation.

It is a matter of controversy whether T2-weighted images or T2*-weighted images are more useful for optimal evaluation of the SPIO-enhanced liver. The results of some previous studies have shown improved detection of hepatic lesions with the use of SPIO-enhanced T2*-weighted spoiled GRE sequences (41,42), but others have reported the opposite result (43). However, the fact that SPIO shortens T2* more than T2 is suggestive of the superiority of SPIO-enhanced T2*-weighted acquisitions (41,42). An echo time (TE) of 6–10 msec usually is sufficient to generate SPIO-enhanced T2*-weighted images of the cirrhotic liver (44).

SPIO-enhanced imaging has a number of disadvantages. First, well-differentiated hepatocellular carcinomas and dysplastic nodules in which some Kupffer cell function is retained may accumulate SPIO particles and escape detection. Second, confluent areas of benign fibrosis, which lack Kupffer cells, may appear masslike and mimic hepatocellular carcinoma. Third, field inhomogeneity induced by SPIO particles in the surrounding liver parenchyma may obscure small hepatocellular carcinomas unless small voxels are used. Fourth, macrovascular invasion may be difficult to detect. A further difficulty is created by the fact that SPIO particles must be infused slowly, typically over 30 minutes, which lengthens the examination time if unenhanced images are desired.

To overcome these limitations of SPIO for diagnostic imaging, as well as the limitations of individual gadolinium chelates, some institutions advocate double contrast-enhanced MR imaging in which both SPIO particles and a gadolinium chelate are administered sequentially in the same examination. The results of initial studies show that such examinations are feasible (45) and that the use of the two contrast agents may increase the conspicuity of hepatic tumors (46). A theoretical advantage of such protocols is the potential synergy of the contrast agents in making hepatocellular carcinomas more obvious, because the SPIO particles cause nonneoplastic liver tissue to appear darker, while the gadolinium chelate causes hepatocellular carcinomas to appear brighter. However, to exploit that synergy, T1 and T2* weighting must be sufficient; in other words, spoiled GRE sequences with a high flip angle and relatively long TE must be used (47). Another predicted advantage is that SPIO and gadolinium provide complementary biologic data that may be used to improve nodule detection and characterization beyond the levels achievable with the use of a single contrast agent (21,47). However, to our knowledge, these hypothetical advantages have not yet been tested empirically in a study comparing double contrast-enhanced imaging with gadolinium-enhanced imaging or SPIO-enhanced imaging performed separately.

A shortcoming of double contrast-enhanced imaging is that it is logistically more difficult than contrast-enhanced imaging with the use of a single agent. It also exposes the patient to two drugs and, thus, to a greater risk of adverse effects (48). A potential technical limitation of SPIO is that, in addition to shortening the T2* of liver tissue, it shortens T1 and therefore reduces the effectiveness of the subsequent gadolinium injection at dynamic T1-weighted imaging. An example of this limitation was reported by Kim et al, who observed that SPIO accumulation in background liver parenchyma impaired the assessment of lesion enhancement at dynamic imaging (49). This subject merits further investigation.

Contrast Agent Safety.— Each category of contrast agents is associated with a characteristic group of possible adverse effects. Adverse events occur in less than 2% of patients after the administration of extracellular gadolinium formulations with a low molecular weight, and most such events are transient and of limited severity. The most commonly reported adverse effects of this group of agents are headache, nausea, and vomiting. Anaphylactoid reactions are rare and typically occur in patients with a history of respiratory difficulties or respiratory allergic reactions (50).

Gadobenate dimeglumine has an adverse effect profile similar to that of extracellular gadolinium formulations (51). Headache is the most common adverse event, but nausea, vomiting, a warm sensation at the injection site, and anaphylactoid reaction also have been reported. Although gadobenate dimeglumine is excreted in part by the liver, no dosage modification is necessary for patients with hepatic insufficiency.

Nephrogenic systemic fibrosis is a recently described potential delayed complication of gadolinium administration in patients with severe renal insufficiency. In May 2007, the Food and Drug Administration issued a warning that the use of gadolinium should be avoided in patients with acute renal insufficiency due to hepatorenal syndrome and in liver transplant candidates in the perioperative period, since such patients may have a higher risk for this adverse effect (52).

The most commonly reported adverse effects of SPIO administration are lumbar pain, anaphylactoid reaction, and hypotension (53,54). SPIO should be used with caution in patients with an endogenous iron overload: A typical SPIO dose for a 70-kg patient (0.56 mg Fe/kg) contains approximately 39 mg of iron, 10% of the iron content in a 2-unit blood transfusion.

Regarding double contrast-enhanced MR imaging, preliminary observations suggest that the use of two contrast agents (SPIO and a gadolinium-based agent) does not increase the frequency of adverse events beyond that associated with the use of one of the agents and that the sequential administration of the contrast agents is safe in most patients (55).

	Differential Diagnosis

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
The MR imaging features that are relevant for the differential diagnosis of cirrhosis-associated hepatocellular nodules are summarized in Table 5 and are correlated in this section of the article with gross pathologic and histologic features.

Regenerative nodules characteristically are surrounded by regions of bridging fibrosis (Fig 1) (2). In the early stages of cirrhosis, the liver may contain nonregenerative nodules of hepatocellular tissue carved out by bridging fibrosis. Such nodules may resemble regenerative nodules at gross pathologic evaluation; however, because they lack regenerative features, they are not classified as regenerative.

View larger version (173K): [in this window] [in a new window] [Download PPT slide]   Figure 1a.  Regenerative nodules in a 54-year-old man with HCV-induced cirrhosis. (a–c) Axial two-dimensional (2D) T2*-weighted spoiled GRE images obtained at 3 T with a TE of 5.8 msec. (a) Unenhanced image shows minimal heterogeneity of the liver parenchyma, with faintly visible nodules of various sizes (arrows). (b) Image obtained after SPIO administration shows a marked loss of signal intensity because of the phagocytic uptake of SPIO particles in the nodules (arrows), which appear dark and sharply circumscribed. (c) Double contrast-enhanced image obtained after the intravenous administration of a gadolinium-based contrast agent shows fibrotic reticuli with high signal intensity due to extracellular accumulation of the low-molecular-weight agent. The enhancement of fibrotic tissue further increases the visibility of the innumerable nodules (arrows). (d) Photograph of explanted liver from a 67-year-old woman with HCV-induced cirrhosis shows an outer surface studded with regenerative nodules of various sizes. (e) Photomicrograph (original magnification, x40; hematoxylin-eosin [H-E] stain) of a slice from the specimen shown in d shows the hyperplastic proliferation of hepatocytes in nodular formations (arrows) surrounded by fibrotic septa (arrowheads).

View larger version (179K): [in this window] [in a new window] [Download PPT slide]   Figure 1b.  Regenerative nodules in a 54-year-old man with HCV-induced cirrhosis. (a–c) Axial two-dimensional (2D) T2*-weighted spoiled GRE images obtained at 3 T with a TE of 5.8 msec. (a) Unenhanced image shows minimal heterogeneity of the liver parenchyma, with faintly visible nodules of various sizes (arrows). (b) Image obtained after SPIO administration shows a marked loss of signal intensity because of the phagocytic uptake of SPIO particles in the nodules (arrows), which appear dark and sharply circumscribed. (c) Double contrast-enhanced image obtained after the intravenous administration of a gadolinium-based contrast agent shows fibrotic reticuli with high signal intensity due to extracellular accumulation of the low-molecular-weight agent. The enhancement of fibrotic tissue further increases the visibility of the innumerable nodules (arrows). (d) Photograph of explanted liver from a 67-year-old woman with HCV-induced cirrhosis shows an outer surface studded with regenerative nodules of various sizes. (e) Photomicrograph (original magnification, x40; hematoxylin-eosin [H-E] stain) of a slice from the specimen shown in d shows the hyperplastic proliferation of hepatocytes in nodular formations (arrows) surrounded by fibrotic septa (arrowheads).

View larger version (182K): [in this window] [in a new window] [Download PPT slide]   Figure 1c.  Regenerative nodules in a 54-year-old man with HCV-induced cirrhosis. (a–c) Axial two-dimensional (2D) T2*-weighted spoiled GRE images obtained at 3 T with a TE of 5.8 msec. (a) Unenhanced image shows minimal heterogeneity of the liver parenchyma, with faintly visible nodules of various sizes (arrows). (b) Image obtained after SPIO administration shows a marked loss of signal intensity because of the phagocytic uptake of SPIO particles in the nodules (arrows), which appear dark and sharply circumscribed. (c) Double contrast-enhanced image obtained after the intravenous administration of a gadolinium-based contrast agent shows fibrotic reticuli with high signal intensity due to extracellular accumulation of the low-molecular-weight agent. The enhancement of fibrotic tissue further increases the visibility of the innumerable nodules (arrows). (d) Photograph of explanted liver from a 67-year-old woman with HCV-induced cirrhosis shows an outer surface studded with regenerative nodules of various sizes. (e) Photomicrograph (original magnification, x40; hematoxylin-eosin [H-E] stain) of a slice from the specimen shown in d shows the hyperplastic proliferation of hepatocytes in nodular formations (arrows) surrounded by fibrotic septa (arrowheads).

View larger version (151K): [in this window] [in a new window] [Download PPT slide]   Figure 1d.  Regenerative nodules in a 54-year-old man with HCV-induced cirrhosis. (a–c) Axial two-dimensional (2D) T2*-weighted spoiled GRE images obtained at 3 T with a TE of 5.8 msec. (a) Unenhanced image shows minimal heterogeneity of the liver parenchyma, with faintly visible nodules of various sizes (arrows). (b) Image obtained after SPIO administration shows a marked loss of signal intensity because of the phagocytic uptake of SPIO particles in the nodules (arrows), which appear dark and sharply circumscribed. (c) Double contrast-enhanced image obtained after the intravenous administration of a gadolinium-based contrast agent shows fibrotic reticuli with high signal intensity due to extracellular accumulation of the low-molecular-weight agent. The enhancement of fibrotic tissue further increases the visibility of the innumerable nodules (arrows). (d) Photograph of explanted liver from a 67-year-old woman with HCV-induced cirrhosis shows an outer surface studded with regenerative nodules of various sizes. (e) Photomicrograph (original magnification, x40; hematoxylin-eosin [H-E] stain) of a slice from the specimen shown in d shows the hyperplastic proliferation of hepatocytes in nodular formations (arrows) surrounded by fibrotic septa (arrowheads).

View larger version (182K): [in this window] [in a new window] [Download PPT slide]   Figure 1e.  Regenerative nodules in a 54-year-old man with HCV-induced cirrhosis. (a–c) Axial two-dimensional (2D) T2*-weighted spoiled GRE images obtained at 3 T with a TE of 5.8 msec. (a) Unenhanced image shows minimal heterogeneity of the liver parenchyma, with faintly visible nodules of various sizes (arrows). (b) Image obtained after SPIO administration shows a marked loss of signal intensity because of the phagocytic uptake of SPIO particles in the nodules (arrows), which appear dark and sharply circumscribed. (c) Double contrast-enhanced image obtained after the intravenous administration of a gadolinium-based contrast agent shows fibrotic reticuli with high signal intensity due to extracellular accumulation of the low-molecular-weight agent. The enhancement of fibrotic tissue further increases the visibility of the innumerable nodules (arrows). (d) Photograph of explanted liver from a 67-year-old woman with HCV-induced cirrhosis shows an outer surface studded with regenerative nodules of various sizes. (e) Photomicrograph (original magnification, x40; hematoxylin-eosin [H-E] stain) of a slice from the specimen shown in d shows the hyperplastic proliferation of hepatocytes in nodular formations (arrows) surrounded by fibrotic septa (arrowheads).

At histologic analysis, regenerative nodules have an intact reticulin framework, a normal vascular profile, and preserved hepatocellular and phagocytic functions (18). Portal tracts are present, but periportal fibrosis and scarring may lead to ductular proliferation and portal vein obliteration. The hepatocytes in regenerative nodules resemble those in the adjacent liver parenchyma, although subtle nuclear pleomorphism may be observed. Mitotic activity usually is absent. Thus, regenerative nodules resemble normal liver tissue but also show features of regeneration such as twinning of cell plates, distortion of plate architecture with curvilinear cell plates at the periphery of nodules, lack of lipofuscin pigment, and increased numbers of binucleate or multinucleate hepatocytes.

Unenhanced MR Imaging Features.— Regenerative nodules are usually innumerable, but they may be difficult to see on radiologic images, depending on the imaging technique used. When they are visible, regenerative nodules appear sharply circumscribed within the liver parenchyma (57). On unenhanced T2- and T2*-weighted images, the nodules typically display low signal intensity; their signal intensity on T1-weighted images is variable (60). Lipid-containing regenerative nodules display signal loss on out-of-phase GRE images and unenhanced asymmetric spin-echo images in comparison with in-phase images. Steatotic regenerative nodules tend to occur in multiples (Fig 2). A single fatty nodule is suggestive of a dysplastic or malignant process (Fig 3).

View larger version (137K): [in this window] [in a new window] [Download PPT slide]   Figure 2a.  Steatotic regenerative nodules in a 54-year-old woman with cirrhosis secondary to fatty liver disease. (a, b) Comparison of axial unenhanced 2D spoiled GRE out-of-phase (TE, 2.3 msec) (a) and in-phase (TE, 4.6 msec) (b) images obtained at 1.5 T shows a diffuse signal intensity loss in the liver in a because of phase interference–induced fat and water signal cancellation (chemical shift of the second kind), a finding indicative of diffuse steatosis. Superimposed on the steatotic background are multiple steatotic nodules (arrows) with a signal intensity that is markedly lower than that of background in a and slightly higher than that of background in b. (c) Fat fraction map derived from a and b by calculating a value for each pixel with the equation F = (SIP – SOP)/2(SIP), where F is the fat fraction, SIP is the signal intensity of a pixel on the in-phase image, and SOP is the signal intensity of the same pixel on the out-of-phase image. The map shows diffuse low signal intensity in the liver, a finding indicative of mild hepatic steatosis (an absence of steatosis would be represented as zero signal in the liver), with foci of slightly higher signal intensity (arrows) indicative of steatotic nodules.

View larger version (134K): [in this window] [in a new window] [Download PPT slide]   Figure 2b.  Steatotic regenerative nodules in a 54-year-old woman with cirrhosis secondary to fatty liver disease. (a, b) Comparison of axial unenhanced 2D spoiled GRE out-of-phase (TE, 2.3 msec) (a) and in-phase (TE, 4.6 msec) (b) images obtained at 1.5 T shows a diffuse signal intensity loss in the liver in a because of phase interference–induced fat and water signal cancellation (chemical shift of the second kind), a finding indicative of diffuse steatosis. Superimposed on the steatotic background are multiple steatotic nodules (arrows) with a signal intensity that is markedly lower than that of background in a and slightly higher than that of background in b. (c) Fat fraction map derived from a and b by calculating a value for each pixel with the equation F = (SIP – SOP)/2(SIP), where F is the fat fraction, SIP is the signal intensity of a pixel on the in-phase image, and SOP is the signal intensity of the same pixel on the out-of-phase image. The map shows diffuse low signal intensity in the liver, a finding indicative of mild hepatic steatosis (an absence of steatosis would be represented as zero signal in the liver), with foci of slightly higher signal intensity (arrows) indicative of steatotic nodules.

View larger version (113K): [in this window] [in a new window] [Download PPT slide]   Figure 2c.  Steatotic regenerative nodules in a 54-year-old woman with cirrhosis secondary to fatty liver disease. (a, b) Comparison of axial unenhanced 2D spoiled GRE out-of-phase (TE, 2.3 msec) (a) and in-phase (TE, 4.6 msec) (b) images obtained at 1.5 T shows a diffuse signal intensity loss in the liver in a because of phase interference–induced fat and water signal cancellation (chemical shift of the second kind), a finding indicative of diffuse steatosis. Superimposed on the steatotic background are multiple steatotic nodules (arrows) with a signal intensity that is markedly lower than that of background in a and slightly higher than that of background in b. (c) Fat fraction map derived from a and b by calculating a value for each pixel with the equation F = (SIP – SOP)/2(SIP), where F is the fat fraction, SIP is the signal intensity of a pixel on the in-phase image, and SOP is the signal intensity of the same pixel on the out-of-phase image. The map shows diffuse low signal intensity in the liver, a finding indicative of mild hepatic steatosis (an absence of steatosis would be represented as zero signal in the liver), with foci of slightly higher signal intensity (arrows) indicative of steatotic nodules.

View larger version (121K): [in this window] [in a new window] [Download PPT slide]   Figure 3a.  Steatotic hepatocellular carcinoma in a 48-year-old man with HCV-induced cirrhosis. (a–c) Axial SPIO-enhanced 2D spoiled GRE images obtained at 1.5 T with TEs of 2.6 msec (a), 4.8 msec (b), and 6.6 msec (c) show a 15-mm nodule in liver segment VIII (arrow). A loss of signal intensity in the nodule periphery on the out-of-phase images (a, c) in comparison with that on the in-phase image (b) is indicative of intralesional fat. (d) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image shows heterogeneous enhancement of the nodule. (e) Photograph of a section of the liver, which was explanted 45 days after MR imaging, shows a yellowish nodule (arrows) in an anatomic location corresponding to that of the nodule in a–d. Histologic analysis showed it to be a moderately differentiated steatotic hepatocellular carcinoma.

View larger version (128K): [in this window] [in a new window] [Download PPT slide]   Figure 3b.  Steatotic hepatocellular carcinoma in a 48-year-old man with HCV-induced cirrhosis. (a–c) Axial SPIO-enhanced 2D spoiled GRE images obtained at 1.5 T with TEs of 2.6 msec (a), 4.8 msec (b), and 6.6 msec (c) show a 15-mm nodule in liver segment VIII (arrow). A loss of signal intensity in the nodule periphery on the out-of-phase images (a, c) in comparison with that on the in-phase image (b) is indicative of intralesional fat. (d) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image shows heterogeneous enhancement of the nodule. (e) Photograph of a section of the liver, which was explanted 45 days after MR imaging, shows a yellowish nodule (arrows) in an anatomic location corresponding to that of the nodule in a–d. Histologic analysis showed it to be a moderately differentiated steatotic hepatocellular carcinoma.

View larger version (131K): [in this window] [in a new window] [Download PPT slide]   Figure 3c.  Steatotic hepatocellular carcinoma in a 48-year-old man with HCV-induced cirrhosis. (a–c) Axial SPIO-enhanced 2D spoiled GRE images obtained at 1.5 T with TEs of 2.6 msec (a), 4.8 msec (b), and 6.6 msec (c) show a 15-mm nodule in liver segment VIII (arrow). A loss of signal intensity in the nodule periphery on the out-of-phase images (a, c) in comparison with that on the in-phase image (b) is indicative of intralesional fat. (d) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image shows heterogeneous enhancement of the nodule. (e) Photograph of a section of the liver, which was explanted 45 days after MR imaging, shows a yellowish nodule (arrows) in an anatomic location corresponding to that of the nodule in a–d. Histologic analysis showed it to be a moderately differentiated steatotic hepatocellular carcinoma.

View larger version (110K): [in this window] [in a new window] [Download PPT slide]   Figure 3d.  Steatotic hepatocellular carcinoma in a 48-year-old man with HCV-induced cirrhosis. (a–c) Axial SPIO-enhanced 2D spoiled GRE images obtained at 1.5 T with TEs of 2.6 msec (a), 4.8 msec (b), and 6.6 msec (c) show a 15-mm nodule in liver segment VIII (arrow). A loss of signal intensity in the nodule periphery on the out-of-phase images (a, c) in comparison with that on the in-phase image (b) is indicative of intralesional fat. (d) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image shows heterogeneous enhancement of the nodule. (e) Photograph of a section of the liver, which was explanted 45 days after MR imaging, shows a yellowish nodule (arrows) in an anatomic location corresponding to that of the nodule in a–d. Histologic analysis showed it to be a moderately differentiated steatotic hepatocellular carcinoma.

View larger version (141K): [in this window] [in a new window] [Download PPT slide]   Figure 3e.  Steatotic hepatocellular carcinoma in a 48-year-old man with HCV-induced cirrhosis. (a–c) Axial SPIO-enhanced 2D spoiled GRE images obtained at 1.5 T with TEs of 2.6 msec (a), 4.8 msec (b), and 6.6 msec (c) show a 15-mm nodule in liver segment VIII (arrow). A loss of signal intensity in the nodule periphery on the out-of-phase images (a, c) in comparison with that on the in-phase image (b) is indicative of intralesional fat. (d) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image shows heterogeneous enhancement of the nodule. (e) Photograph of a section of the liver, which was explanted 45 days after MR imaging, shows a yellowish nodule (arrows) in an anatomic location corresponding to that of the nodule in a–d. Histologic analysis showed it to be a moderately differentiated steatotic hepatocellular carcinoma.

Contrast-enhanced MR Imaging Features.— Contrast-enhanced imaging features are diagnostically more specific than findings at unenhanced imaging. After the injection of an extracellular gadolinium chelate, most regenerative nodules enhance to the same degree as the adjacent liver or show slightly less enhancement. Uptake and excretion of gadobenate dimeglumine by these nodules is usually preserved. Consequently, on images acquired during the hepatocellular phase after an injection of gadobenate dimeglumine, virtually all regenerative nodules have a similar signal intensity, which gives the liver a homogeneous appearance (Fig 4). Occasionally, regenerative nodules may have sufficient hepatocellular function to take up the hepatocellular agent but not to excrete it; such nodules show hyperintense signal (24,38). Finally, because most regenerative nodules have a preserved phagocytic function, they are SPIO avid and appear hypointense on SPIO-enhanced T2- and T2*-weighted images (Fig 1).

View larger version (139K): [in this window] [in a new window] [Download PPT slide]   Figure 4a.  Regenerative nodules and hepatocellular carcinoma in a 57-year-old man with cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows a small hepatocellular carcinoma (arrow), which has higher signal intensity than the regenerative nodules surrounding it, because of its lesser phagocytic uptake of SPIO particles. (b f) Axial 3D fat-saturated T1-weighted spoiled GRE images acquired at 3 T 1 week later, before (b) and in three phases after (c f) an intravenous injection of a gadolinium-based contrast agent. The unenhanced image (b) does not show the small hepatocellular carcinoma, but the hepatic arterial phase image (c) shows increased signal intensity in the carcinoma (arrow). (d, e) Portal venous (d) and equilibrium (e) phase images show washout in the lesion center, which has lower signal intensity than that of the liver parenchyma, while contrast material retained in the lesion rim results in hyperintense signal suggestive of a capsule or pseudocapsule. The innumerable regenerative nodules have signal intensity that is varied in b, isointense to that of the parenchyma in c, and slightly less intense than that of the parenchyma in e. (f) Hepatocellular phase image obtained 90 minutes after the injection shows slightly lower signal intensity in the carcinoma because of less active hepatocellular function than in background liver. Because most of the regenerative nodules have taken up some gadolinium, the cirrhotic parenchyma appears more homogeneous than in e. However, the regenerative nodules and fibrous reticuli are most visible in a, the SPIO-enhanced image.

View larger version (117K): [in this window] [in a new window] [Download PPT slide]   Figure 4b.  Regenerative nodules and hepatocellular carcinoma in a 57-year-old man with cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows a small hepatocellular carcinoma (arrow), which has higher signal intensity than the regenerative nodules surrounding it, because of its lesser phagocytic uptake of SPIO particles. (b f) Axial 3D fat-saturated T1-weighted spoiled GRE images acquired at 3 T 1 week later, before (b) and in three phases after (c f) an intravenous injection of a gadolinium-based contrast agent. The unenhanced image (b) does not show the small hepatocellular carcinoma, but the hepatic arterial phase image (c) shows increased signal intensity in the carcinoma (arrow). (d, e) Portal venous (d) and equilibrium (e) phase images show washout in the lesion center, which has lower signal intensity than that of the liver parenchyma, while contrast material retained in the lesion rim results in hyperintense signal suggestive of a capsule or pseudocapsule. The innumerable regenerative nodules have signal intensity that is varied in b, isointense to that of the parenchyma in c, and slightly less intense than that of the parenchyma in e. (f) Hepatocellular phase image obtained 90 minutes after the injection shows slightly lower signal intensity in the carcinoma because of less active hepatocellular function than in background liver. Because most of the regenerative nodules have taken up some gadolinium, the cirrhotic parenchyma appears more homogeneous than in e. However, the regenerative nodules and fibrous reticuli are most visible in a, the SPIO-enhanced image.

View larger version (117K): [in this window] [in a new window] [Download PPT slide]   Figure 4c.  Regenerative nodules and hepatocellular carcinoma in a 57-year-old man with cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows a small hepatocellular carcinoma (arrow), which has higher signal intensity than the regenerative nodules surrounding it, because of its lesser phagocytic uptake of SPIO particles. (b f) Axial 3D fat-saturated T1-weighted spoiled GRE images acquired at 3 T 1 week later, before (b) and in three phases after (c f) an intravenous injection of a gadolinium-based contrast agent. The unenhanced image (b) does not show the small hepatocellular carcinoma, but the hepatic arterial phase image (c) shows increased signal intensity in the carcinoma (arrow). (d, e) Portal venous (d) and equilibrium (e) phase images show washout in the lesion center, which has lower signal intensity than that of the liver parenchyma, while contrast material retained in the lesion rim results in hyperintense signal suggestive of a capsule or pseudocapsule. The innumerable regenerative nodules have signal intensity that is varied in b, isointense to that of the parenchyma in c, and slightly less intense than that of the parenchyma in e. (f) Hepatocellular phase image obtained 90 minutes after the injection shows slightly lower signal intensity in the carcinoma because of less active hepatocellular function than in background liver. Because most of the regenerative nodules have taken up some gadolinium, the cirrhotic parenchyma appears more homogeneous than in e. However, the regenerative nodules and fibrous reticuli are most visible in a, the SPIO-enhanced image.

View larger version (130K): [in this window] [in a new window] [Download PPT slide]   Figure 4d.  Regenerative nodules and hepatocellular carcinoma in a 57-year-old man with cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows a small hepatocellular carcinoma (arrow), which has higher signal intensity than the regenerative nodules surrounding it, because of its lesser phagocytic uptake of SPIO particles. (b f) Axial 3D fat-saturated T1-weighted spoiled GRE images acquired at 3 T 1 week later, before (b) and in three phases after (c f) an intravenous injection of a gadolinium-based contrast agent. The unenhanced image (b) does not show the small hepatocellular carcinoma, but the hepatic arterial phase image (c) shows increased signal intensity in the carcinoma (arrow). (d, e) Portal venous (d) and equilibrium (e) phase images show washout in the lesion center, which has lower signal intensity than that of the liver parenchyma, while contrast material retained in the lesion rim results in hyperintense signal suggestive of a capsule or pseudocapsule. The innumerable regenerative nodules have signal intensity that is varied in b, isointense to that of the parenchyma in c, and slightly less intense than that of the parenchyma in e. (f) Hepatocellular phase image obtained 90 minutes after the injection shows slightly lower signal intensity in the carcinoma because of less active hepatocellular function than in background liver. Because most of the regenerative nodules have taken up some gadolinium, the cirrhotic parenchyma appears more homogeneous than in e. However, the regenerative nodules and fibrous reticuli are most visible in a, the SPIO-enhanced image.

View larger version (121K): [in this window] [in a new window] [Download PPT slide]   Figure 4e.  Regenerative nodules and hepatocellular carcinoma in a 57-year-old man with cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows a small hepatocellular carcinoma (arrow), which has higher signal intensity than the regenerative nodules surrounding it, because of its lesser phagocytic uptake of SPIO particles. (b f) Axial 3D fat-saturated T1-weighted spoiled GRE images acquired at 3 T 1 week later, before (b) and in three phases after (c f) an intravenous injection of a gadolinium-based contrast agent. The unenhanced image (b) does not show the small hepatocellular carcinoma, but the hepatic arterial phase image (c) shows increased signal intensity in the carcinoma (arrow). (d, e) Portal venous (d) and equilibrium (e) phase images show washout in the lesion center, which has lower signal intensity than that of the liver parenchyma, while contrast material retained in the lesion rim results in hyperintense signal suggestive of a capsule or pseudocapsule. The innumerable regenerative nodules have signal intensity that is varied in b, isointense to that of the parenchyma in c, and slightly less intense than that of the parenchyma in e. (f) Hepatocellular phase image obtained 90 minutes after the injection shows slightly lower signal intensity in the carcinoma because of less active hepatocellular function than in background liver. Because most of the regenerative nodules have taken up some gadolinium, the cirrhotic parenchyma appears more homogeneous than in e. However, the regenerative nodules and fibrous reticuli are most visible in a, the SPIO-enhanced image.

View larger version (135K): [in this window] [in a new window] [Download PPT slide]   Figure 4f.  Regenerative nodules and hepatocellular carcinoma in a 57-year-old man with cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows a small hepatocellular carcinoma (arrow), which has higher signal intensity than the regenerative nodules surrounding it, because of its lesser phagocytic uptake of SPIO particles. (b f) Axial 3D fat-saturated T1-weighted spoiled GRE images acquired at 3 T 1 week later, before (b) and in three phases after (c f) an intravenous injection of a gadolinium-based contrast agent. The unenhanced image (b) does not show the small hepatocellular carcinoma, but the hepatic arterial phase image (c) shows increased signal intensity in the carcinoma (arrow). (d, e) Portal venous (d) and equilibrium (e) phase images show washout in the lesion center, which has lower signal intensity than that of the liver parenchyma, while contrast material retained in the lesion rim results in hyperintense signal suggestive of a capsule or pseudocapsule. The innumerable regenerative nodules have signal intensity that is varied in b, isointense to that of the parenchyma in c, and slightly less intense than that of the parenchyma in e. (f) Hepatocellular phase image obtained 90 minutes after the injection shows slightly lower signal intensity in the carcinoma because of less active hepatocellular function than in background liver. Because most of the regenerative nodules have taken up some gadolinium, the cirrhotic parenchyma appears more homogeneous than in e. However, the regenerative nodules and fibrous reticuli are most visible in a, the SPIO-enhanced image.

Regenerative nodules with a diameter of more than 15 mm at imaging have an increased likelihood of being dysplastic or malignant. However, an absence of early enhancement after gadolinium administration, preserved uptake of gadobenate dimeglumine, and avid SPIO accumulation are findings suggestive of benignity.

Siderotic Nodules
Because of their high iron content, siderotic nodules have low signal intensity on T1- and T2*-weighted unenhanced MR images (10,58,61,62). These lesions may be either regenerative or dysplastic, and no unenhanced imaging feature (size, number, distribution) permits reliable differentiation between the two (10). To our knowledge, no studies have yet been performed to assess whether contrast-enhanced imaging helps distinguish regenerative siderotic nodules from dysplastic ones.

Dysplastic Nodules
Gross Pathologic and Histologic Features.— In gross appearance, dysplastic nodules may be indistinguishable from regenerative nodules (2). Furthermore, only 15%–28% of cirrhotic liver explants were found to contain dysplastic nodules (10,63–67). The detection of dysplastic nodules is even more difficult when they are embedded in livers with innumerable regenerative nodules.

Dysplastic nodules are characterized histologically by progressive architectural derangement, nuclear crowding, atypia, and a variable number of unpaired arterioles or capillaries. Low-grade dysplastic nodules closely resemble regenerative nodules histologically: They have a preserved hepatic architecture and minimal cytologic atypia, as well as normal vascular profile, hepatocellular function, and Kupffer cell density (2). They are considered to have low malignant potential with slow, infrequent progression to hepatocellular carcinoma (16).

High-grade dysplastic nodules show some architectural distortion and more advanced cytologic atypia, with sinusoidal "capillarization" and an increased density of unpaired arteries (20) (Fig 5). The Kupffer cell density is variable; it may be increased, normal, or diminished (26–28). Hepatocyte plates are two to three cells wide. High-grade dysplastic nodules are believed to progress to hepatocellular carcinoma more frequently than low-grade dysplastic nodules (16). The high-grade nodules closely resemble well-differentiated hepatocellular carcinomas and are difficult to distinguish histologically, particularly those that are small (5).

View larger version (169K): [in this window] [in a new window] [Download PPT slide]   Figure 5a.  Dysplastic nodule in a 45-year-old woman with alcoholic cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a 10-mm nodule in liver segment V (arrow) that has a higher signal intensity than the surrounding parenchyma because of less intranodular uptake of SPIO. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows nodular enhancement (arrow). (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows a well-defined 12-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) shows a well-defined transition between the liver parenchyma (arrowheads) and nodule (arrows). Increased cellular and capillary density, a higher nuclear-cytoplasmic ratio, and moderate architectural distortion in the nodule are indicative of high-grade dysplasia.

View larger version (129K): [in this window] [in a new window] [Download PPT slide]   Figure 5b.  Dysplastic nodule in a 45-year-old woman with alcoholic cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a 10-mm nodule in liver segment V (arrow) that has a higher signal intensity than the surrounding parenchyma because of less intranodular uptake of SPIO. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows nodular enhancement (arrow). (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows a well-defined 12-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) shows a well-defined transition between the liver parenchyma (arrowheads) and nodule (arrows). Increased cellular and capillary density, a higher nuclear-cytoplasmic ratio, and moderate architectural distortion in the nodule are indicative of high-grade dysplasia.

View larger version (145K): [in this window] [in a new window] [Download PPT slide]   Figure 5c.  Dysplastic nodule in a 45-year-old woman with alcoholic cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a 10-mm nodule in liver segment V (arrow) that has a higher signal intensity than the surrounding parenchyma because of less intranodular uptake of SPIO. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows nodular enhancement (arrow). (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows a well-defined 12-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) shows a well-defined transition between the liver parenchyma (arrowheads) and nodule (arrows). Increased cellular and capillary density, a higher nuclear-cytoplasmic ratio, and moderate architectural distortion in the nodule are indicative of high-grade dysplasia.

View larger version (165K): [in this window] [in a new window] [Download PPT slide]   Figure 5d.  Dysplastic nodule in a 45-year-old woman with alcoholic cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a 10-mm nodule in liver segment V (arrow) that has a higher signal intensity than the surrounding parenchyma because of less intranodular uptake of SPIO. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows nodular enhancement (arrow). (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows a well-defined 12-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) shows a well-defined transition between the liver parenchyma (arrowheads) and nodule (arrows). Increased cellular and capillary density, a higher nuclear-cytoplasmic ratio, and moderate architectural distortion in the nodule are indicative of high-grade dysplasia.

Unenhanced MR Imaging Features.— As expected on the basis of their histopathologic characteristics, dysplastic nodules have variable appearances on MR images, and their signal intensity characteristics overlap with those of regenerative nodules and well-differentiated hepatocellular carcinomas. On T2-weighted images, low-grade dysplastic nodules tend to have low signal intensity relative to adjacent liver, whereas high-grade dysplastic nodules tend to have slightly higher signal intensity. T1-weighted images are not helpful because both low- and high-grade dysplastic nodules display variable (low, intermediate, or high) signal intensity.

Contrast-enhanced MR Imaging Features.— On gadolinium- and SPIO-enhanced images, low-grade dysplastic nodules typically are indistinguishable from regenerative nodules, whereas high-grade dysplastic nodules are indistinguishable from well-differentiated hepatocellular carcinomas (Fig 5) (2,26). It is not yet clear whether gadobenate dimeglumine–enhanced imaging permits the characterization of dysplastic nodules.

Hepatocellular Carcinomas
Gross Pathologic and Histologic Features.— At gross examination, hepatocellular carcinomas may be of any size and usually are firm to the touch. Because of intralesional steatosis, cholestasis, hemorrhage, and lipofuscin, their color typically differs from that of the surrounding liver parenchyma, making them readily identifiable at gross pathologic analysis. However, small hepatocellular carcinomas may be invisible at gross specimen analysis and detectable only with microscopy. Hepatocellular carcinomas that arise within dysplastic nodules are particularly difficult to detect at gross examination. Large hepatocellular carcinomas tend to show evidence of necrosis (Fig 6) (2) and often have a mosaic appearance characterized by a seemingly random distribution of confluent small nodules with intervening fibrous septa and areas of necrosis. Tumor capsules, irregular margins, satellite nodules, and vascular extension are frequent findings in such cases (2,32,60).

View larger version (177K): [in this window] [in a new window] [Download PPT slide]   Figure 6a.  Poorly differentiated hepatocellular carcinoma in a 58-year-old man with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE = 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly defined high-signal-intensity mass (arrows) at the junction of liver segments V and VIII, abutting the liver capsule and causing it to retract. (b, c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE images obtained in hepatic arterial (b) and equilibrium (c) phases show heterogeneous enhancement of the mass, with rimlike (white arrowheads) and geographic (black arrowheads) high-signal-intensity areas and a low-signal-intensity center (arrow), a finding suggestive of ischemia or necrosis. (d) Photograph of a gross specimen resected 3 months later shows an infiltrative mass (arrows) in a location corresponding to that in a–c, with prominent areas of desmoplastic reaction (*), hemorrhage, and necrosis. A definitive diagnosis was made at microscopy.

View larger version (129K): [in this window] [in a new window] [Download PPT slide]   Figure 6b.  Poorly differentiated hepatocellular carcinoma in a 58-year-old man with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE = 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly defined high-signal-intensity mass (arrows) at the junction of liver segments V and VIII, abutting the liver capsule and causing it to retract. (b, c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE images obtained in hepatic arterial (b) and equilibrium (c) phases show heterogeneous enhancement of the mass, with rimlike (white arrowheads) and geographic (black arrowheads) high-signal-intensity areas and a low-signal-intensity center (arrow), a finding suggestive of ischemia or necrosis. (d) Photograph of a gross specimen resected 3 months later shows an infiltrative mass (arrows) in a location corresponding to that in a–c, with prominent areas of desmoplastic reaction (*), hemorrhage, and necrosis. A definitive diagnosis was made at microscopy.

View larger version (151K): [in this window] [in a new window] [Download PPT slide]   Figure 6c.  Poorly differentiated hepatocellular carcinoma in a 58-year-old man with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE = 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly defined high-signal-intensity mass (arrows) at the junction of liver segments V and VIII, abutting the liver capsule and causing it to retract. (b, c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE images obtained in hepatic arterial (b) and equilibrium (c) phases show heterogeneous enhancement of the mass, with rimlike (white arrowheads) and geographic (black arrowheads) high-signal-intensity areas and a low-signal-intensity center (arrow), a finding suggestive of ischemia or necrosis. (d) Photograph of a gross specimen resected 3 months later shows an infiltrative mass (arrows) in a location corresponding to that in a–c, with prominent areas of desmoplastic reaction (*), hemorrhage, and necrosis. A definitive diagnosis was made at microscopy.

View larger version (108K): [in this window] [in a new window] [Download PPT slide]   Figure 6d.  Poorly differentiated hepatocellular carcinoma in a 58-year-old man with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE = 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly defined high-signal-intensity mass (arrows) at the junction of liver segments V and VIII, abutting the liver capsule and causing it to retract. (b, c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE images obtained in hepatic arterial (b) and equilibrium (c) phases show heterogeneous enhancement of the mass, with rimlike (white arrowheads) and geographic (black arrowheads) high-signal-intensity areas and a low-signal-intensity center (arrow), a finding suggestive of ischemia or necrosis. (d) Photograph of a gross specimen resected 3 months later shows an infiltrative mass (arrows) in a location corresponding to that in a–c, with prominent areas of desmoplastic reaction (*), hemorrhage, and necrosis. A definitive diagnosis was made at microscopy.

Histologic features of hepatocellular carcinoma include advanced architectural distortion (widening and irregularity of hepatocyte plates, presence of pseudoglandular structures, absence of portal tracts, and increased density of singleton arteries), nuclear atypia, necrosis, and microscopic invasion of stroma and portal tracts. Microscopic invasion is the primary diagnostic criterion for distinguishing well-differentiated hepatocellular carcinomas from dysplastic nodules (2). A tumor capsule composed of an inner layer of fibrous tissue and an outer layer of compressed vessels and bile ducts is evident at histologic examination in 65%–82% of large hepatocellular carcinomas, but that finding is not required for a diagnosis of hepatocellular carcinoma. A capsule also may be present in regenerative nodules and dysplastic nodules. Small hepatocellular carcinomas tend to be well differentiated (Fig 7). Large hepatocellular carcinomas are most often moderately (Fig 8) or poorly differentiated (Fig 6) (2). The presence of extracapsular extension or macrovascular invasion, the absence of a tumor capsule, and poor histologic differentiation are associated with a higher risk of tumor recurrence after treatment.

View larger version (140K): [in this window] [in a new window] [Download PPT slide]   Figure 7a.  Well-differentiated hypovascular hepatocellular carcinoma in a 55-year-old woman with alcoholic and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows an 18-mm nodule (arrow) in liver segment II with relative signal hyperintensity because of diminished SPIO accumulation. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows faint enhancement of the nodule (arrow). (c) Photograph of a cut surface of the liver, which was explanted 2 months later, shows a well-defined 18-mm pale nodule (arrows) in a location corresponding to that in a and b. A fibrous septum divides the nodule into anterior and posterior compartments. (d) Photomicrograph (original magnification, x20; H-E stain) of a specimen from the nodule (arrowheads) shows moderate architectural distortion and augmented cellular density. Microscopic analysis at a higher magnification showed increased mitotic activity and an absence of necrosis and hemorrhage, characteristic features of well-differentiated hepatocellular carcinoma.

View larger version (124K): [in this window] [in a new window] [Download PPT slide]   Figure 7b.  Well-differentiated hypovascular hepatocellular carcinoma in a 55-year-old woman with alcoholic and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows an 18-mm nodule (arrow) in liver segment II with relative signal hyperintensity because of diminished SPIO accumulation. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows faint enhancement of the nodule (arrow). (c) Photograph of a cut surface of the liver, which was explanted 2 months later, shows a well-defined 18-mm pale nodule (arrows) in a location corresponding to that in a and b. A fibrous septum divides the nodule into anterior and posterior compartments. (d) Photomicrograph (original magnification, x20; H-E stain) of a specimen from the nodule (arrowheads) shows moderate architectural distortion and augmented cellular density. Microscopic analysis at a higher magnification showed increased mitotic activity and an absence of necrosis and hemorrhage, characteristic features of well-differentiated hepatocellular carcinoma.

View larger version (131K): [in this window] [in a new window] [Download PPT slide]   Figure 7c.  Well-differentiated hypovascular hepatocellular carcinoma in a 55-year-old woman with alcoholic and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows an 18-mm nodule (arrow) in liver segment II with relative signal hyperintensity because of diminished SPIO accumulation. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows faint enhancement of the nodule (arrow). (c) Photograph of a cut surface of the liver, which was explanted 2 months later, shows a well-defined 18-mm pale nodule (arrows) in a location corresponding to that in a and b. A fibrous septum divides the nodule into anterior and posterior compartments. (d) Photomicrograph (original magnification, x20; H-E stain) of a specimen from the nodule (arrowheads) shows moderate architectural distortion and augmented cellular density. Microscopic analysis at a higher magnification showed increased mitotic activity and an absence of necrosis and hemorrhage, characteristic features of well-differentiated hepatocellular carcinoma.

View larger version (192K): [in this window] [in a new window] [Download PPT slide]   Figure 7d.  Well-differentiated hypovascular hepatocellular carcinoma in a 55-year-old woman with alcoholic and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows an 18-mm nodule (arrow) in liver segment II with relative signal hyperintensity because of diminished SPIO accumulation. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the arterial phase shows faint enhancement of the nodule (arrow). (c) Photograph of a cut surface of the liver, which was explanted 2 months later, shows a well-defined 18-mm pale nodule (arrows) in a location corresponding to that in a and b. A fibrous septum divides the nodule into anterior and posterior compartments. (d) Photomicrograph (original magnification, x20; H-E stain) of a specimen from the nodule (arrowheads) shows moderate architectural distortion and augmented cellular density. Microscopic analysis at a higher magnification showed increased mitotic activity and an absence of necrosis and hemorrhage, characteristic features of well-differentiated hepatocellular carcinoma.

View larger version (130K): [in this window] [in a new window] [Download PPT slide]   Figure 8a.  Moderately differentiated hepatocellular carcinoma in a 62-year-old man with alcoholic cirrhosis. (a) Axial 2D SPIO-enhanced T2*-weighted (TE = 6.6 msec) spoiled GRE image shows a 25-mm nodule (arrowheads) in liver segment VIII. Signal intensity in the nodule is higher than that in the surrounding liver parenchyma because of diminished SPIO uptake. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows heterogeneous enhancement of the nodule. (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows an encapsulated 28-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) of a slice from the nodule center shows moderate cellular atypia with an increased nuclear-cytoplasmic ratio (arrowheads) and disorganized blood vessels (arrows). Microvascular invasion and spread of tumor cells to the adjacent parenchyma, features characteristic of moderate differentiation, also were found.

View larger version (115K): [in this window] [in a new window] [Download PPT slide]   Figure 8b.  Moderately differentiated hepatocellular carcinoma in a 62-year-old man with alcoholic cirrhosis. (a) Axial 2D SPIO-enhanced T2*-weighted (TE = 6.6 msec) spoiled GRE image shows a 25-mm nodule (arrowheads) in liver segment VIII. Signal intensity in the nodule is higher than that in the surrounding liver parenchyma because of diminished SPIO uptake. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows heterogeneous enhancement of the nodule. (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows an encapsulated 28-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) of a slice from the nodule center shows moderate cellular atypia with an increased nuclear-cytoplasmic ratio (arrowheads) and disorganized blood vessels (arrows). Microvascular invasion and spread of tumor cells to the adjacent parenchyma, features characteristic of moderate differentiation, also were found.

View larger version (137K): [in this window] [in a new window] [Download PPT slide]   Figure 8c.  Moderately differentiated hepatocellular carcinoma in a 62-year-old man with alcoholic cirrhosis. (a) Axial 2D SPIO-enhanced T2*-weighted (TE = 6.6 msec) spoiled GRE image shows a 25-mm nodule (arrowheads) in liver segment VIII. Signal intensity in the nodule is higher than that in the surrounding liver parenchyma because of diminished SPIO uptake. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows heterogeneous enhancement of the nodule. (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows an encapsulated 28-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) of a slice from the nodule center shows moderate cellular atypia with an increased nuclear-cytoplasmic ratio (arrowheads) and disorganized blood vessels (arrows). Microvascular invasion and spread of tumor cells to the adjacent parenchyma, features characteristic of moderate differentiation, also were found.

View larger version (174K): [in this window] [in a new window] [Download PPT slide]   Figure 8d.  Moderately differentiated hepatocellular carcinoma in a 62-year-old man with alcoholic cirrhosis. (a) Axial 2D SPIO-enhanced T2*-weighted (TE = 6.6 msec) spoiled GRE image shows a 25-mm nodule (arrowheads) in liver segment VIII. Signal intensity in the nodule is higher than that in the surrounding liver parenchyma because of diminished SPIO uptake. (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows heterogeneous enhancement of the nodule. (c) Photograph of a gross pathologic section of the liver, which was explanted 2 months later, shows an encapsulated 28-mm nodule (arrows) in a location corresponding to that in a and b. (d) Photomicrograph (original magnification, x100; H-E stain) of a slice from the nodule center shows moderate cellular atypia with an increased nuclear-cytoplasmic ratio (arrowheads) and disorganized blood vessels (arrows). Microvascular invasion and spread of tumor cells to the adjacent parenchyma, features characteristic of moderate differentiation, also were found.

Unenhanced MR Imaging Features.— The signal intensity characteristics of hepatocellular carcinomas depend on their size, grade, and biologic features. On T1-weighted images, small hepatocellular carcinomas may have variable signal intensity; on T2-weighted images, signal intensity usually is slightly increased (60). Some well-differentiated hepatocellular carcinomas may appear isointense or even hypointense in comparison with the surrounding liver parenchyma.

Large hepatocellular carcinomas exhibit a greater array of signal intensity alterations caused by intralesional fat, hemorrhage, and necrosis than is seen in smaller lesions. Steatotic hepatocellular carcinomas are characterized by a signal intensity decrease on out-of-phase images in comparison with in-phase images (Fig 3), and they are easily identified by comparing fat-saturated images with non–fat-saturated ones. Hemorrhagic hepatocellular carcinomas may have marked high signal intensity on T1-weighted images and low signal intensity on T2- and T2*-weighted images. Intralesional necrosis typically manifests as one or more areas of low signal intensity on T1-weighted images and high signal intensity on T2-weighted images.

Contrast-enhanced MR Imaging Features.— Approximately 80%–90% of hepatocellular carcinomas are hypervascular and become intensely enhanced during the arterial phase after a bolus injection of a gadolinium chelate (Fig 4). In delayed imaging phases, hypervascular hepatocellular carcinomas undergo washout and typically have signal intensity lower than that of the surrounding liver; however, some may appear isointense to the liver parenchyma and therefore may be difficult to see on images obtained during the portal venous phase and later phases.

About 10%–20% of hepatocellular carcinomas are hypovascular and show contrast enhancement slightly less than that in the surrounding liver on arterial phase images. Typically, hypovascular hepatocellular carcinomas are small, well-differentiated tumors. However, poorly differentiated and diffusely infiltrating hypovascular hepatocellular carcinomas also may occur (24,39). Such lesions may be difficult to detect on gadolinium-enhanced MR images despite their large size and aggressive behavior, but they are usually visible on SPIO-enhanced images. Hepatocellular carcinomas also may occur as multiple diffuse contiguous nodules.

Because of their diminished uptake of gadolinium, hepatocellular carcinomas appear as unenhanced foci of low signal intensity in comparison with the surrounding nonneoplastic liver parenchyma on hepatocellular phase images (23,24) (Fig 4).

Moderately and poorly differentiated hepatocellular carcinomas characteristically accumulate less SPIO than the surrounding liver parenchyma and have relatively high signal intensity on T2- and T2*-weighted SPIO-enhanced images. Well-differentiated hepatocellular carcinomas may accumulate SPIO and tend to be iso- or hypointense compared with the background liver. Thus, the degree of SPIO uptake may be used as a noninvasive means of grading hepatocellular carcinomas (26–28). Large hepatocellular carcinomas may have nonuniform Kupffer cell density and show heterogeneous accumulation of SPIO particles.

Sequential administration of SPIO and gadolinium in the same MR imaging session permits the assessment of both vascularity and phagocytic function. When SPIO is administered first, the findings on SPIO-enhanced images are those described earlier. However, the appearance on subsequent dynamic gadolinium-enhanced images may be altered, and, as discussed previously, the assessment of venous washout may be impaired. Thus, the manner in which dynamic images are interpreted in double contrast-enhanced protocols may require some modification. However, to our knowledge, this matter has not yet been investigated.

Structural Variation of Large Hepatocellular Carcinomas.— Small hepatocellular carcinomas tend to be homogeneous, round or oval, and well defined. By contrast, large hepatocellular carcinomas may exhibit a broad spectrum of morphologic features, including a mosaic pattern, a tumor capsule, an intratumoral nodule ("nodule-in-nodule" appearance), and extracapsular extension with the formation of satellite nodules.

The mosaic pattern reflects underlying mosaic pathologic features and is defined by the presence of multiple compartments of variable signal intensity at unenhanced T1-, T2-, and T2*-weighted imaging (60). The compartments are distributed within the mass in a seemingly random manner and enhance to variable degrees after the administration of contrast material (Fig 9).

View larger version (144K): [in this window] [in a new window] [Download PPT slide]   Figure 9a.  Hepatocellular carcinoma with a mosaic pattern of enhancement in a 62-year-old man with HCV-induced cirrhosis. (a) Axial unenhanced 3D fat-saturated T1-weighted spoiled GRE image obtained at 3 T shows a well-circumscribed mass (arrow) at the junction of liver segments VII and VIII with signal isointense to that of the liver parenchyma but with a lower-signal-intensity periphery (arrowheads) that represents a capsule. (b–d) Axial gadolinium-enhanced images obtained with the same sequence at 3 T during the hepatic arterial (b), portal venous (c), and equilibrium (d) phases show heterogeneous enhancement of the mass (white arrow), with randomly distributed areas of hyper- and hypoenhancement (mosaic pattern). A satellite nodule (black arrow) with signal that is hyperintense in b but progressively less intense in c and d also is visible. (e) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows high signal intensity in the mass (white arrow) and satellite nodule (black arrow) because of a lack of SPIO uptake in comparison with that in the liver parenchyma.

View larger version (159K): [in this window] [in a new window] [Download PPT slide]   Figure 9b.  Hepatocellular carcinoma with a mosaic pattern of enhancement in a 62-year-old man with HCV-induced cirrhosis. (a) Axial unenhanced 3D fat-saturated T1-weighted spoiled GRE image obtained at 3 T shows a well-circumscribed mass (arrow) at the junction of liver segments VII and VIII with signal isointense to that of the liver parenchyma but with a lower-signal-intensity periphery (arrowheads) that represents a capsule. (b–d) Axial gadolinium-enhanced images obtained with the same sequence at 3 T during the hepatic arterial (b), portal venous (c), and equilibrium (d) phases show heterogeneous enhancement of the mass (white arrow), with randomly distributed areas of hyper- and hypoenhancement (mosaic pattern). A satellite nodule (black arrow) with signal that is hyperintense in b but progressively less intense in c and d also is visible. (e) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows high signal intensity in the mass (white arrow) and satellite nodule (black arrow) because of a lack of SPIO uptake in comparison with that in the liver parenchyma.

View larger version (186K): [in this window] [in a new window] [Download PPT slide]   Figure 9c.  Hepatocellular carcinoma with a mosaic pattern of enhancement in a 62-year-old man with HCV-induced cirrhosis. (a) Axial unenhanced 3D fat-saturated T1-weighted spoiled GRE image obtained at 3 T shows a well-circumscribed mass (arrow) at the junction of liver segments VII and VIII with signal isointense to that of the liver parenchyma but with a lower-signal-intensity periphery (arrowheads) that represents a capsule. (b–d) Axial gadolinium-enhanced images obtained with the same sequence at 3 T during the hepatic arterial (b), portal venous (c), and equilibrium (d) phases show heterogeneous enhancement of the mass (white arrow), with randomly distributed areas of hyper- and hypoenhancement (mosaic pattern). A satellite nodule (black arrow) with signal that is hyperintense in b but progressively less intense in c and d also is visible. (e) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows high signal intensity in the mass (white arrow) and satellite nodule (black arrow) because of a lack of SPIO uptake in comparison with that in the liver parenchyma.

View larger version (184K): [in this window] [in a new window] [Download PPT slide]   Figure 9d.  Hepatocellular carcinoma with a mosaic pattern of enhancement in a 62-year-old man with HCV-induced cirrhosis. (a) Axial unenhanced 3D fat-saturated T1-weighted spoiled GRE image obtained at 3 T shows a well-circumscribed mass (arrow) at the junction of liver segments VII and VIII with signal isointense to that of the liver parenchyma but with a lower-signal-intensity periphery (arrowheads) that represents a capsule. (b–d) Axial gadolinium-enhanced images obtained with the same sequence at 3 T during the hepatic arterial (b), portal venous (c), and equilibrium (d) phases show heterogeneous enhancement of the mass (white arrow), with randomly distributed areas of hyper- and hypoenhancement (mosaic pattern). A satellite nodule (black arrow) with signal that is hyperintense in b but progressively less intense in c and d also is visible. (e) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows high signal intensity in the mass (white arrow) and satellite nodule (black arrow) because of a lack of SPIO uptake in comparison with that in the liver parenchyma.

View larger version (179K): [in this window] [in a new window] [Download PPT slide]   Figure 9e.  Hepatocellular carcinoma with a mosaic pattern of enhancement in a 62-year-old man with HCV-induced cirrhosis. (a) Axial unenhanced 3D fat-saturated T1-weighted spoiled GRE image obtained at 3 T shows a well-circumscribed mass (arrow) at the junction of liver segments VII and VIII with signal isointense to that of the liver parenchyma but with a lower-signal-intensity periphery (arrowheads) that represents a capsule. (b–d) Axial gadolinium-enhanced images obtained with the same sequence at 3 T during the hepatic arterial (b), portal venous (c), and equilibrium (d) phases show heterogeneous enhancement of the mass (white arrow), with randomly distributed areas of hyper- and hypoenhancement (mosaic pattern). A satellite nodule (black arrow) with signal that is hyperintense in b but progressively less intense in c and d also is visible. (e) Axial SPIO-enhanced 2D T2*-weighted (TE, 5.8 msec) spoiled GRE image shows high signal intensity in the mass (white arrow) and satellite nodule (black arrow) because of a lack of SPIO uptake in comparison with that in the liver parenchyma.

A tumor capsule appears as a thin circumferential rim at the periphery of a hepatocellular carcinoma nodule on MR images. It typically thickens with increasing nodule size (68). On unenhanced T1- and T2-weighted images, the tumor capsule often has low signal intensity; however, a capsule thicker than 4 mm may have an outer layer with high signal intensity on T2-weighted images. Because capsules are composed of fibrous tissue and compressed vessels, they enhance progressively after the administration of gadolinium and retain contrast enhancement longer, showing signal intensity higher than that of the surrounding parenchyma on delayed phase images (Fig 4) (68–70). Tumor capsules do not accumulate SPIO particles, and they have high signal intensity on SPIO-enhanced T2-weighted and T2*-weighted images. Although a capsule is characteristic of hepatocellular carcinomas, capsules also may be present in large regenerative nodules (Fig 10) and dysplastic nodules.

View larger version (146K): [in this window] [in a new window] [Download PPT slide]   Figure 10a.  Large regenerative nodule with capsule in a 46-year-old woman with HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows innumerable SPIO-avid nodules (*) in liver segment VII. The largest nodule (arrows) has a diameter of 14 mm. (b) Axial double contrast-enhanced 2D spoiled GRE image obtained with a delay after administration of gadolinium in the same imaging session as a shows enhanced signal intensity in the rim of the largest nodule (arrowheads), a finding suggestive of a capsule. (c) Photograph of a section of the explanted liver shows a 25-mm nodule (arrow) with a well-defined capsule (arrowheads). At microscopic analysis, it proved to be a regenerative macronodule with no atypia.

View larger version (145K): [in this window] [in a new window] [Download PPT slide]   Figure 10b.  Large regenerative nodule with capsule in a 46-year-old woman with HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows innumerable SPIO-avid nodules (*) in liver segment VII. The largest nodule (arrows) has a diameter of 14 mm. (b) Axial double contrast-enhanced 2D spoiled GRE image obtained with a delay after administration of gadolinium in the same imaging session as a shows enhanced signal intensity in the rim of the largest nodule (arrowheads), a finding suggestive of a capsule. (c) Photograph of a section of the explanted liver shows a 25-mm nodule (arrow) with a well-defined capsule (arrowheads). At microscopic analysis, it proved to be a regenerative macronodule with no atypia.

View larger version (131K): [in this window] [in a new window] [Download PPT slide]   Figure 10c.  Large regenerative nodule with capsule in a 46-year-old woman with HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows innumerable SPIO-avid nodules (*) in liver segment VII. The largest nodule (arrows) has a diameter of 14 mm. (b) Axial double contrast-enhanced 2D spoiled GRE image obtained with a delay after administration of gadolinium in the same imaging session as a shows enhanced signal intensity in the rim of the largest nodule (arrowheads), a finding suggestive of a capsule. (c) Photograph of a section of the explanted liver shows a 25-mm nodule (arrow) with a well-defined capsule (arrowheads). At microscopic analysis, it proved to be a regenerative macronodule with no atypia.

A characteristic but uncommon finding is the nodule-in-nodule appearance, which generally represents a hepatocellular carcinoma within a larger dysplastic or regenerative nodule. On arterial phase images obtained after gadolinium administration, the carcinomatous nodule typically appears enhanced, while the surrounding regenerative or dysplastic nodule does not enhance and has lower signal intensity. On SPIO-enhanced T2- and T2*-weighted images, signal in the carcinoma usually appears hyperintense, while that in the regenerative nodule or dysplastic nodule is hypointense (Fig 5) (67,71,72). At dynamic gadolinium-enhanced imaging, necrotic hepatocellular carcinomas that contain nodules of viable tumor tissue also may have a nodule-in-nodule appearance (Fig 11).

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 11a.  Nodule-in-nodule appearance of hepatocellular carcinoma in a 56-year-old woman with HCV-induced cirrhosis. (a, b) Axial SPIO-enhanced 2D T2*-weighted spoiled GRE (TE, 6.6 msec) (a) and T2-weighted echo-train spin-echo (TE, 90 msec) (b) images obtained at 1.5 T show a 35-mm heterogeneously hyperintense mass in the right liver lobe (arrows). (c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows a markedly enhanced mural nodule (arrowheads) within the largely nonenhanced mass. The lack of enhancement in the mass is suggestive of necrosis. (d) Photograph of a section from the liver, which was explanted 10 weeks after MR imaging, shows two mural nodules (arrows) within the mass. At histologic analysis, the nodules proved to be viable hepatocellular carcinoma; the rest of the mass consisted of necrotic tissue. Note the multiple regenerative nodules (*) of various sizes in the surrounding liver parenchyma.

View larger version (145K): [in this window] [in a new window] [Download PPT slide]   Figure 11b.  Nodule-in-nodule appearance of hepatocellular carcinoma in a 56-year-old woman with HCV-induced cirrhosis. (a, b) Axial SPIO-enhanced 2D T2*-weighted spoiled GRE (TE, 6.6 msec) (a) and T2-weighted echo-train spin-echo (TE, 90 msec) (b) images obtained at 1.5 T show a 35-mm heterogeneously hyperintense mass in the right liver lobe (arrows). (c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows a markedly enhanced mural nodule (arrowheads) within the largely nonenhanced mass. The lack of enhancement in the mass is suggestive of necrosis. (d) Photograph of a section from the liver, which was explanted 10 weeks after MR imaging, shows two mural nodules (arrows) within the mass. At histologic analysis, the nodules proved to be viable hepatocellular carcinoma; the rest of the mass consisted of necrotic tissue. Note the multiple regenerative nodules (*) of various sizes in the surrounding liver parenchyma.

View larger version (159K): [in this window] [in a new window] [Download PPT slide]   Figure 11c.  Nodule-in-nodule appearance of hepatocellular carcinoma in a 56-year-old woman with HCV-induced cirrhosis. (a, b) Axial SPIO-enhanced 2D T2*-weighted spoiled GRE (TE, 6.6 msec) (a) and T2-weighted echo-train spin-echo (TE, 90 msec) (b) images obtained at 1.5 T show a 35-mm heterogeneously hyperintense mass in the right liver lobe (arrows). (c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows a markedly enhanced mural nodule (arrowheads) within the largely nonenhanced mass. The lack of enhancement in the mass is suggestive of necrosis. (d) Photograph of a section from the liver, which was explanted 10 weeks after MR imaging, shows two mural nodules (arrows) within the mass. At histologic analysis, the nodules proved to be viable hepatocellular carcinoma; the rest of the mass consisted of necrotic tissue. Note the multiple regenerative nodules (*) of various sizes in the surrounding liver parenchyma.

View larger version (144K): [in this window] [in a new window] [Download PPT slide]   Figure 11d.  Nodule-in-nodule appearance of hepatocellular carcinoma in a 56-year-old woman with HCV-induced cirrhosis. (a, b) Axial SPIO-enhanced 2D T2*-weighted spoiled GRE (TE, 6.6 msec) (a) and T2-weighted echo-train spin-echo (TE, 90 msec) (b) images obtained at 1.5 T show a 35-mm heterogeneously hyperintense mass in the right liver lobe (arrows). (c) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows a markedly enhanced mural nodule (arrowheads) within the largely nonenhanced mass. The lack of enhancement in the mass is suggestive of necrosis. (d) Photograph of a section from the liver, which was explanted 10 weeks after MR imaging, shows two mural nodules (arrows) within the mass. At histologic analysis, the nodules proved to be viable hepatocellular carcinoma; the rest of the mass consisted of necrotic tissue. Note the multiple regenerative nodules (*) of various sizes in the surrounding liver parenchyma.

Extracapsular extension with the formation of satellite lesions is frequently seen in large nodular hepatocellular carcinomas. The satellite lesions often appear as multiple subcentimeter nodules outside the tumor margins (Fig 12). Advanced hepatocellular carcinomas may penetrate the liver capsule and invade adjacent structures such as the diaphragm and abdominal wall.

View larger version (154K): [in this window] [in a new window] [Download PPT slide]   Figure 12a.  Hepatocellular carcinoma with satellite nodules in a 67-year-old woman with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly circumscribed mass (arrow) in liver segment VI with multiple subcentimeter peripheral nodules (arrowheads). (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows avid enhancement of the lesions (arrows). (c) Photograph of a gross section from the explanted liver shows a 2-cm dominant nodule (arrows) and multiple subcentimeter satellite nodules (*). At microscopic analysis, the nodules proved to be moderately differentiated hepatocellular carcinoma.

View larger version (114K): [in this window] [in a new window] [Download PPT slide]   Figure 12b.  Hepatocellular carcinoma with satellite nodules in a 67-year-old woman with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly circumscribed mass (arrow) in liver segment VI with multiple subcentimeter peripheral nodules (arrowheads). (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows avid enhancement of the lesions (arrows). (c) Photograph of a gross section from the explanted liver shows a 2-cm dominant nodule (arrows) and multiple subcentimeter satellite nodules (*). At microscopic analysis, the nodules proved to be moderately differentiated hepatocellular carcinoma.

View larger version (139K): [in this window] [in a new window] [Download PPT slide]   Figure 12c.  Hepatocellular carcinoma with satellite nodules in a 67-year-old woman with HBV- and HCV-induced cirrhosis. (a) Axial SPIO-enhanced 2D T2*-weighted (TE, 6.6 msec) spoiled GRE image obtained at 1.5 T shows a poorly circumscribed mass (arrow) in liver segment VI with multiple subcentimeter peripheral nodules (arrowheads). (b) Axial gadolinium-enhanced 3D fat-saturated T1-weighted spoiled GRE image obtained during the hepatic arterial phase shows avid enhancement of the lesions (arrows). (c) Photograph of a gross section from the explanted liver shows a 2-cm dominant nodule (arrows) and multiple subcentimeter satellite nodules (*). At microscopic analysis, the nodules proved to be moderately differentiated hepatocellular carcinoma.

Another characteristic of large hepatocellular carcinomas is macrovascular invasion of the portal and hepatic veins (73). On unenhanced and SPIO-enhanced images, vascular invasion may be difficult to detect, although a disturbance in the expected flow-related signal intensity alterations may be suggestive of the diagnosis. The diagnosis usually is straightforward at gadolinium-enhanced imaging: The intravascular tumor tissue rapidly enhances during the arterial phase, and a filling defect is seen on images acquired during later phases (32,60). A solid tumor within the vascular lumen is commonly referred to as a tumor thrombus. A tumor thrombus enhances early and expands the vascular lumen, whereas a nonneoplastic bland thrombus does not enhance and, instead of expanding the lumen, causes it to contract. Differentiation of the two types of thrombi is critical. A tumor thrombus, by definition, is indicative of hepatocellular carcinoma and may be an important clue for the diagnosis of diffuse infiltrating cancer that otherwise might be mistaken for cirrhosis-induced parenchymal heterogeneity or confluent areas of fibrosis. The presence of a tumor thrombus also conveys a high risk of hematogenous dissemination of cancer and precludes liver transplantation as a treatment option. By comparison, a bland thrombus is a frequent finding in the setting of cirrhosis, may occur in the absence of hepatocellular carcinoma, and, depending on its location and extent, may be of minimal importance for decision making with regard to disease management.

	Summary

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 
Cirrhosis-associated hepatocellular nodules are a heterogeneous group of hepatic lesions that are classified histologically either as regenerative (benign) or as dysplastic or neoplastic (premalignant or malignant). However, the accurate detection and characterization of these nodules may be difficult even at histopathologic analysis. Differential diagnosis may be facilitated by comparing the clinical and pathologic findings with MR imaging features. Although different categories of hepatocellular nodules have imaging characteristics that overlap considerably, a comprehensive familiarity with diagnostically specific features at unenhanced and contrast-enhanced MR imaging may help distinguish benign lesions from premalignant and malignant ones.

	Footnotes

 

Abbreviations: GRE = gradient echo, HBV = hepatitis B virus, HCV = hepatitis C virus, H-E = hematoxylin-eosin, SPIO = superparamagnetic iron oxide, TE = echo time, 3D = three-dimensional, 2D = two-dimensional

² Current address: Department of Radiology, Fundación Santa Fe de Bogotá University Hospital, Bogotá, Colombia.

	References

Top Abstract LEARNING OBJECTIVES Introduction Nodule Classification Nodule Characterization Differential Diagnosis Summary References

 

1. Popper H. Pathologic aspects of cirrhosis: a review. Am J Pathol 1977;87:228–264.[Medline]
2. Terminology of nodular hepatocellular lesions. International Working Party. Hepatology 1995;22: 983–993.[CrossRef][Medline]
3. Coleman WB. Mechanisms of human hepatocarcinogenesis. Curr Mol Med 2003;3:573–588.[CrossRef][Medline]
4. Wurmbach E, Chen YB, Khitrov G, et al. Genome-wide molecular profiles of HCV-induced dysplasia and hepatocellular carcinoma. Hepatology 2007; 45:938–947.[CrossRef][Medline]
5. Bruix J, Sherman M. Management of hepatocellular carcinoma. Hepatology 2005;42:1208–1236.[CrossRef][Medline]
6. United Network for Organ Sharing. Policy 3.6, Organ Distribution: Allocation of Livers. Organ Procurement and Transplantation Network Web site. http://www.optn.org/policiesAndBylaws/policies.asp. Revised September 18, 2007. Accessed March 3, 2008.
7. Edmondson HA, Steiner PE. Primary carcinoma of the liver: a study of 100 cases among 48,900 necropsies. Cancer 1954;7:462–503.[CrossRef][Medline]
8. Yuki K, Hirohashi S, Sakamoto M, Kanai T, Shimosato Y. Growth and spread of hepatocellular carcinoma: a review of 240 consecutive autopsy cases. Cancer 1990;66:2174–2179.[CrossRef][Medline]
9. Kanai T, Hirohashi S, Upton MP, et al. Pathology of small hepatocellular carcinoma: a proposal for a new gross classification. Cancer 1987;60:810–819.[CrossRef][Medline]
10. Krinsky GA, Lee VS, Nguyen MT, et al. Siderotic nodules at MR imaging: regenerative or dysplastic? J Comput Assist Tomogr 2000;24:773–776.[CrossRef][Medline]
11. Henmi A, Uchida T, Shikata T. Karyometric analysis of liver cell dysplasia and hepatocellular carcinoma: evidence against precancerous nature of liver cell dysplasia. Cancer 1985;55:2594–2599.[Medline]
12. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet 2003;362:1907–1917.[CrossRef][Medline]
13. Efremidis SC, Hytiroglou P. The multistep process of hepatocarcinogenesis in cirrhosis with imaging correlation. Eur Radiol 2002;12:753–764.[CrossRef][Medline]
14. Wang XW, Hussain SP, Huo TI, et al. Molecular pathogenesis of human hepatocellular carcinoma. Toxicology 2002;181–182:43–47.
15. Staib F, Hussain SP, Hofseth LJ, Wang XW, Harris CC. TP53 and liver carcinogenesis. Hum Mutat 2003;21:201–216.[CrossRef][Medline]
16. Borzio M, Fargion S, Borzio F, et al. Impact of large regenerative, low grade and high grade dysplastic nodules in hepatocellular carcinoma development. J Hepatol 2003;39:208–214.[CrossRef][Medline]
17. Roncalli M. Hepatocellular nodules in cirrhosis: focus on diagnostic criteria on liver biopsy—a Western experience. Liver Transpl 2004;10:S9–S15.[CrossRef][Medline]
18. Roncalli M, Roz E, Coggi G, et al. The vascular profile of regenerative and dysplastic nodules of the cirrhotic liver: implications for diagnosis and classification. Hepatology 1999;30:1174–1178.[CrossRef][Medline]
19. Choi BI, Han JK, Hong SH, et al. Dysplastic nodules of the liver: imaging findings. Abdom Imaging 1999;24:250–257.[CrossRef][Medline]
20. Park YN, Yang CP, Fernandez GJ, Cubukcu O, Thung SN, Theise ND. Neoangiogenesis and sinusoidal "capillarization" in dysplastic nodules of the liver. Am J Surg Pathol 1998;22:656–662.[CrossRef][Medline]
21. Bhartia B, Ward J, Guthrie JA, Robinson PJ. Hepatocellular carcinoma in cirrhotic livers: double-contrast thin-section MR imaging with pathologic correlation of explanted tissue. AJR Am J Roentgenol 2003;180:577–584.[Abstract/Free Full Text]
22. Gandhi SN, Brown MA, Wong JG, Aguirre DA, Sirlin CB. MR contrast agents for liver imaging: what, when, how. RadioGraphics 2006;26:1621–1636.[Abstract/Free Full Text]
23. Tanimoto A, Kuwatsuru R, Kadoya M, et al. Evaluation of gadobenate dimeglumine in hepatocellular carcinoma: results from phase II and phase III clinical trials in Japan. J Magn Reson Imaging 1999;10:450–460.[CrossRef][Medline]
24. Vogl TJ, Stupavsky A, Pegios W, et al. Hepatocellular carcinoma: evaluation with dynamic and static gadobenate dimeglumine–enhanced MR imaging and histopathologic correlation. Radiology 1997; 205:721–728.[Abstract/Free Full Text]
25. Grazioli L, Morana G, Caudana R, et al. Hepatocellular carcinoma: correlation between gadobenate dimeglumine-enhanced MRI and pathologic findings. Invest Radiol 2000;35:25–34.[CrossRef][Medline]
26. Lim JH, Choi D, Cho SK, et al. Conspicuity of hepatocellular nodular lesions in cirrhotic livers at ferumoxides-enhanced MR imaging: importance of Kupffer cell number. Radiology 2001;220:669– 676.[Abstract/Free Full Text]
27. Imai Y, Murakami T, Yoshida S, et al. Superparamagnetic iron oxide-enhanced magnetic resonance images of hepatocellular carcinoma: correlation with histological grading. Hepatology 2000;32: 205–212.[CrossRef][Medline]
28. Asahina Y, Izumi N, Uchihara M, et al. Assessment of Kupffer cells by ferumoxides-enhanced MR imaging is beneficial for diagnosis of hepatocellular carcinoma: comparison of pathological diagnosis and perfusion patterns assessed by CT hepatic arteriography and CT arterioportography. Hepatol Res 2003;27:196–204.[CrossRef][Medline]
29. Cui R, He J, Zhang F, et al. Diagnostic value of protein induced by vitamin K absence (PIVKAII) and hepatoma-specific band of serum gamma-glutamyl transferase (GGTII) as hepatocellular carcinoma markers complementary to alpha-fetoprotein. Br J Cancer 2003;88:1878–1882.[CrossRef][Medline]
30. Sawabu N, Wakabayashi T, Ozaki K, et al. Serum tumor markers in patients with hepatocellular carcinoma: diagnosis of alpha-fetoprotein–low or –negative patients. Gastroenterol Jpn 1985;20: 209–215.[Medline]
31. Tsai JF, Jeng JE, Chuang LY, et al. Serum insulin-like growth factor-II and alpha-fetoprotein as tumor markers of hepatocellular carcinoma. Tumour Biol 2003;24:291–298.[CrossRef][Medline]
32. Hussain SM, Semelka RC, Mitchell DG. MR imaging of hepatocellular carcinoma. Magn Reson Imaging Clin N Am 2002;10:31–52, v.[CrossRef][Medline]
33. Taouli B, Vilgrain V, Dumont E, Daire JL, Fan B, Menu Y. Evaluation of liver diffusion isotropy and characterization of focal hepatic lesions with two single-shot echo-planar MR imaging sequences: prospective study in 66 patients. Radiology 2003; 226:71–78.[Abstract/Free Full Text]
34. Yamashita Y, Mitsuzaki K, Yi T, et al. Small hepatocellular carcinoma in patients with chronic liver damage: prospective comparison of detection with dynamic MR imaging and helical CT of the whole liver. Radiology 1996;200:79–84.[Abstract/Free Full Text]
35. Oi H, Murakami T, Kim T, Matsushita M, Kishimoto H, Nakamura H. Dynamic MR imaging and early-phase helical CT for detecting small intrahepatic metastases of hepatocellular carcinoma. AJR Am J Roentgenol 1996;166:369–374.[Abstract/Free Full Text]
36. Ros PR, Freeny PC, Harms SE, et al. Hepatic MR imaging with ferumoxides: a multicenter clinical trial of the safety and efficacy in the detection of focal hepatic lesions. Radiology 1995;196:481–488.[Abstract/Free Full Text]
37. Stark DD, Weissleder R, Elizondo G, et al. Superparamagnetic iron oxide: clinical application as a contrast agent for MR imaging of the liver. Radiology 1988;168:297–301.[Abstract/Free Full Text]
38. Manfredi R, Maresca G, Baron RL, et al. Gadobenate dimeglumine (BOPTA) enhanced MR imaging: patterns of enhancement in normal liver and cirrhosis. J Magn Reson Imaging 1998;8:862–867.[Medline]
39. Lutz AM, Willmann JK, Goepfert K, Marincek B, Weishaupt D. Hepatocellular carcinoma in cirrhosis: enhancement patterns at dynamic gadolinium- and superparamagnetic iron oxide–enhanced T1-weighted MR imaging. Radiology 2005;237: 520–528.[Abstract/Free Full Text]
40. Qayyum A, Graser A, Westphalen A, et al. CT of benign hypervascular liver nodules in autoimmune hepatitis. AJR Am J Roentgenol 2004;183: 1573–1576.[Abstract/Free Full Text]
41. Tang Y, Yamashita Y, Arakawa A, et al. Detection of hepatocellular carcinoma arising in cirrhotic livers: comparison of gadolinium- and ferumoxides-enhanced MR imaging. AJR Am J Roentgenol 1999; 172:1547–1554.[Abstract/Free Full Text]
42. Arbab AS, Ichikawa T, Araki T, et al. Detection of hepatocellular carcinoma and its metastases with various pulse sequences using superparamagnetic iron oxide (SHU-555-A). Abdom Imaging 2000; 25:151–158.[CrossRef][Medline]
43. Goshima S, Kanematsu M, Matsuo M, et al. Malignant hepatic tumor detection with ferumoxide-enhanced magnetic resonance imaging: is chemical-shift-selective fat suppression necessary for fast spin-echo sequence? J Magn Reson Imaging 2004; 20:75–82.[CrossRef][Medline]
44. Mori K, Scheidler J, Helmberger T, et al. Detection of malignant hepatic lesions before orthotopic liver transplantation: accuracy of ferumoxides-enhanced MR imaging. AJR Am J Roentgenol 2002; 179:1045–1051.[Abstract/Free Full Text]
45. Kubaska S, Sahani DV, Saini S, Hahn PF, Halpern E. Dual contrast enhanced magnetic resonance imaging of the liver with superparamagnetic iron oxide followed by gadolinium for lesion detection and characterization. Clin Radiol 2001;56:410–415.[CrossRef][Medline]
46. Weissleder R, Saini S, Stark DD, Wittenberg J, Ferrucci JT. Dual-contrast MR imaging of liver cancer in rats. AJR Am J Roentgenol 1988;150:561–566.[Abstract/Free Full Text]
47. 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]
48. Kim YK, Kim CS, Lee YH, Kwak HS, Lee JM. Comparison of superparamagnetic iron oxide-enhanced and gadobenate dimeglumine-enhanced dynamic MRI for detection of small hepatocellular carcinomas. AJR Am J Roentgenol 2004;182: 1217–1223.[Abstract/Free Full Text]
49. Kim MJ, Kim JH, Chung JJ, Park MS, Lim JS, Oh YT. Focal hepatic lesions: detection and characterization with combination gadolinium- and superparamagnetic iron oxide–enhanced MR imaging. Radiology 2003;228:719–726.[Abstract/Free Full Text]
50. Bellin MF. MR contrast agents, the old and the new. Eur J Radiol 2006;60:314–323.[CrossRef][Medline]
51. Knopp MV, Bourne MW, Sardanelli F, et al. Gadobenate dimeglumine-enhanced MRI of the breast: analysis of dose response and comparison with gadopentetate dimeglumine. AJR Am J Roentgenol 2003;181:663–676.[Abstract/Free Full Text]
52. United States Food and Drug Administration. Information for healthcare professionals: gadolinium-based contrast agents for magnetic resonance imaging (marketed as Magnevist M, Omniscan, OptiMARK, ProHance). FDA Web site. http://www.fda.gov/cder/drug/infopage/gcca/default.htm. Revised May 23, 2007. Accessed June 11, 2007.
53. Fretz CJ, Elizondo G, Weissleder R, Hahn PF, Stark DD, Ferrucci JT Jr. Superparamagnetic iron oxide–enhanced MR imaging: pulse sequence optimization for detection of liver cancer. Radiology 1989;172:393–397.[Abstract/Free Full Text]
54. Runge VM. Safety of approved MR contrast media for intravenous injection. J Magn Reson Imaging 2000;12:205–213.[CrossRef][Medline]
55. Halavaara J, Tervahartiala P, Isoniemi H, Hockerstedt K. Efficacy of sequential use of superparamagnetic iron oxide and gadolinium in liver MR imaging. Acta Radiol 2002;43:180–185.[CrossRef][Medline]
56. Kanematsu M, Hoshi H, Yamada T, et al. Small hepatic nodules in cirrhosis: ultrasonographic, CT, and MR imaging findings. Abdom Imaging 1999; 24:47–55.[CrossRef][Medline]
57. Zhang J, Krinsky GA. Iron-containing nodules of cirrhosis. NMR Biomed 2004;17:459–464.[CrossRef][Medline]
58. Ohtomo K, Itai Y, Ohtomo Y, Shiga J, Iio M. Regenerating nodules of liver cirrhosis: MR imaging with pathologic correlation. AJR Am J Roentgenol 1990;154:505–507.[Abstract/Free Full Text]
59. Brancatelli G, Federle MP, Grazioli L, Golfieri R, Lencioni R. Large regenerative nodules in Budd-Chiari syndrome and other vascular disorders of the liver: CT and MR imaging findings with clinicopathologic correlation. AJR Am J Roentgenol 2002;178:877–883.[Abstract/Free Full Text]
60. Hussain SM, Zondervan PE, IJzermans JN, Schalm SW, de Man RA, Krestin GP. Benign versus malignant hepatic nodules: MR imaging findings with pathologic correlation. RadioGraphics 2002;22:1023–1036; discussion 1037–1039.[Abstract/Free Full Text]
61. Itai Y, Ohnishi S, Ohtomo K, et al. Regenerating nodules of liver cirrhosis: MR imaging. Radiology 1987;165:419–423.[Abstract/Free Full Text]
62. Murakami T, Kuroda C, Marukawa T, et al. Regenerating nodules in hepatic cirrhosis: MR findings with pathologic correlation. AJR Am J Roentgenol 1990;155:1227–1231.[Abstract/Free Full Text]
63. Hytiroglou P, Theise ND, Schwartz M, Mor E, Miller C, Thung SN. Macroregenerative nodules in a series of adult cirrhotic liver explants: issues of classification and nomenclature. Hepatology 1995; 21:703–708.[Medline]
64. Theise ND, Schwartz M, Miller C, Thung SN. Macroregenerative nodules and hepatocellular carcinoma in forty-four sequential adult liver explants with cirrhosis. Hepatology 1992;16:949–955.[CrossRef][Medline]
65. Ferrell L, Wright T, Lake J, Roberts J, Ascher N. Incidence and diagnostic features of macroregenerative nodules vs. small hepatocellular carcinoma in cirrhotic livers. Hepatology 1992;16:1372–1381.[Medline]
66. Terada T, Nakanuma Y. Survey of iron-accumulative macroregenerative nodules in cirrhotic livers. Hepatology 1989;10:851–854.[CrossRef][Medline]
67. Mitchell DG, Rubin R, Siegelman ES, Burk DL Jr, Rifkin MD. Hepatocellular carcinoma within siderotic regenerative nodules: appearance as a nodule within a nodule on MR images. Radiology 1991;178:101–103.[Abstract/Free Full Text]
68. Kadoya M, Matsui O, Takashima T, Nonomura A. Hepatocellular carcinoma: correlation of MR imaging and histopathologic findings. Radiology 1992;183:819–825.[Abstract/Free Full Text]
69. Itai Y. Capsule of hepatocellular carcinoma: where and how does the capsule show enhancement? Radiology 1999;210:577–579.[Free Full Text]
70. Ishizaki M, Ashida K, Higashi T, et al. The formation of capsule and septum in human hepatocellular carcinoma. Virchows Arch 2001;438:574–580.[CrossRef][Medline]
71. Goshima S, Kanematsu M, Matsuo M, et al. Nodule-in-nodule appearance of hepatocellular carcinomas: comparison of gadolinium-enhanced and ferumoxides-enhanced magnetic resonance imaging. J Magn Reson Imaging 2004;20:250–255.[CrossRef][Medline]
72. Kojiro M. ‘Nodule-in-nodule’ appearance in hepatocellular carcinoma: its significance as a morphologic marker of dedifferentiation. Intervirology 2004;47:179–183.[CrossRef][Medline]
73. Pawlik TM, Poon RT, Abdalla EK, et al. Hepatectomy for hepatocellular carcinoma with major portal or hepatic vein invasion: results of a multicenter study. Surgery 2005;137:403–410.[CrossRef][Medline]

This Article Abstract Figures Only Full Text (PDF) CME Test (opens in a new window) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hanna, R. F. Articles by Sirlin, C. B. Search for Related Content PubMed PubMed Citation Articles by Hanna, R. F. Articles by Sirlin, C. B. Related Collections Magnetic Resonance Imaging Gastrointestinal Radiology