Subscribe to RSS
DOI: 10.1055/s-2001-15497
Imaging Evaluation of the Cirrhotic Liver
Publication History
Publication Date:
31 December 2001 (online)
- ABSTRACT
- PATHOLOGIC AND GROSS CHANGES IN CIRRHOSIS
- IMAGING FEATURE OF CIRRHOSIS
- IMAGING FEATURES OF NODULAR LESIONS IN CIRRHOSIS
- ROLE OF SONOGRAPHIC AND CT SCREENING OF CIRRHOTIC PATIENTS
- SUMMARY
- ABBREVIATIONS
- REFERENCES
ABSTRACT
Because recent advances in medical care decrease the mortality rate due to liver cirrhosis itself, many cirrhotic patients die due to hepatocellular carcinoma. Accordingly, the role of radiology in the evaluation of the patient with cirrhosis is primarily to characterize the morphologic manifestations of the disease, evaluate the hepatic and extrahepatic vasculature, assess the effects of portal hypertension, and detect hepatic tumors. When the latter are identified, a critical role of imaging technology is to differentiate hepatocellular carcinoma from other nodular lesions, such as dysplastic nodules and regenerating nodules.
Screening strategies for patients with cirrhosis have been proposed to facilitate the detection of small, asymptomatic hepatocellular carcinomas. Dynamic studies using computed tomography (CT) and magnetic resonance imaging (MRI) are very useful for the diagnosis of hepatic tumors previously detected by ultrasound, as well as for screening. In Japan, patients with documented cirrhosis typically undergo serum alpha-fetoprotein testing and/or PIVKA-II (protein induced by vitamin K absence or antagonist II) measurements every 2 months, ultrasound every 3 months, and CT or MRI every 6 months. This has resulted in great success in detecting small hepatocellular carcinomas (less than 2 cm in diameter) and early-stage well-differentiated hepatocellular carcinomas.
KEYWORD
Dysplastic nodule - hepatocellular carcinoma - liver cirrhosis - regenerating nodules - screening
Cirrhosis is the ultimate result of alcoholic liver disease, primary biliary cirrhosis, primary sclerosing cholangitis, congestive hepatopathy (cardiac cirrhosis), Wilson's disease, hemochromatosis, and other genetic disorders, and autoimmune diseases, as well as a consequence of chronic infection by hepatitis viruses B, C, and D. In North America, approximately 75% of cases of cirrhosis have been attributed to chronic alcoholism.[1] Although the numbers with viral hepatitis and other identifiable causes have been increasing recently, they have traditionally accounted for about 15%. In contrast, in Asia and Africa, cirrhosis is associated predominantly with chronic viral infection. In Japan hepatitis B is the cause of 17.4% of cases of cirrhosis, hepatitis C of 60.9%, and a combination of hepatitis B and C of 3.8%. Approximately 10% of patients with acute hepatitis B progress to the chronic form of the disease,[2] and about one third of these patients develop potentially progressive disease that can lead to cirrhosis, typically over 2 to 5 years. Hepatitis C has a greater propensity to become chronic and leads to cirrhosis in a higher percentage of patients. It has been estimated that 20% to as many as 50% of patients with hepatitis C may eventually develop cirrhosis,[3] although recent data from Western countries increasingly suggest a figure of 20 to 30% over the first 30 years after initial infection (see Seminars in Liver Disease, vol. 20, issues 1 and 2).
Liver cirrhosis is usually initially suspected on the basis of abnormalities in standard liver function and other biochemical (blood) tests, including viral antibody and antigen testing. Diagnostic imaging is useful for visualizing the morphological and pathological changes in these diffuse liver diseases and may also contribute to clinical diagnostic and therapeutic procedures, for example, image-guided biopsy, tumor ablation therapies, and transjugular intrahepatic portosystemic shunt procedures.
The incidence of hepatocellular carcinoma (HCC) in the United States, once uncommon, is now rising. As in Asia, a major cause of the increase is the epidemic of chronic hepatitis C virus infection. As for chronic hepatitis B virus infection, the relative risk of HCC in patients with cirrhosis due to hepatitis C is approximately 100 times the risk for patients with cirrhosis who are not infected, whereas patients with cirrhosis resulting from alcohol abuse or primary biliary cirrhosis have only a two- to fivefold increased risk of HCC.[4] Cirrhosis due to hepatitis C causes 70% of current cases of HCC in Japan and 30 to 50% in the United States. Early and accurate diagnosis of HCC is important for growing numbers of patients.
The role of radiology in the evaluation of cirrhosis is primarily to characterize the morphologic manifestations of the disease, evaluate the hepatic and extrahepatic vasculature, assess the effects of portal hypertension, and detect hepatic tumors, accurately differentiating HCC from other kinds of tumor. To facilitate the detection of small asymptomatic HCC, screening strategies for patients with cirrhotic liver have been proposed.[5] [6]
#PATHOLOGIC AND GROSS CHANGES IN CIRRHOSIS
Liver cirrhosis is caused by diffuse fibrosis and regenerating nodules that result from liver cell necrosis and degeneration. Nonreversible fibrosis occurs in the Glisson sheath, which destroys the histological structure of the liver.[7] Thus it is characterized by architectural distortion and the development of a spectrum of nodules, ranging from benign regenerating nodules to HCC.[8]
In cirrhosis, in its early stages, the liver may appear normal. With progression of the disease, nodularity of the liver surface and generalized heterogeneity of the hepatic parenchyma can be seen. The porta hepatis and interlobar fissure frequently appear widened due to shrinkage of the right lobe and the medial segment of the left lobe, with concomitant enlargement of the caudate lobe and the lateral segment of the left lobe. These alterations in hepatic morphology tend to be readily apparent on visual inspection of the images.
The gross morphologic appearance of the cirrhotic liver is categorized by the size of the parenchymal nodules: micronodular, macronodular, or mixed.[9] Micronodular cirrhosis is characterized by regenerative nodules of relatively uniform small size. This pattern is seen in chronic alcoholic, hepatitis C, and biliary cirrhosis, and others. In macronodular cirrhosis, the parenchymal nodules are larger, coarser, and more variable in size. The most common cause of macronodular cirrhosis is chronic hepatitis B.
#IMAGING FEATURE OF CIRRHOSIS
Ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI) can be used to visualize the nodular contour of the liver, as in atrophy of the right lobe and hypertrophy of the left and caudate lobes of the liver (Fig. [1]).[10]
#Regenerating Nodules
The regenerating nodules of liver cirrhosis include the macronodular and micronodular types. Some regenerating nodules may show iron deposition, whereas others do not. Regenerating nodules 0.5 to 1.5 cm in size are visible with 3.5- or 5-MHz transducer of ultrasonography; 7.5-MHz transducer can depict them.[11] However, it is sometimes difficult to differentiate cirrhosis from liver tumor because if the regenerating nodule grows to some size, it is visualized with relatively low echo in comparison with the surrounding liver parenchyma, which shows a relatively high echo due to fibrosis.
The regenerating nodules are rarely seen on CT because of limited spatial and contrast resolution. If they contain substantial iron deposition (siderotic regenerating nodules) and are more than 2 to 3 cm in size, they may be recognized as high attenuation nodules onplain CT, or the liver parenchyma including the siderotic regenerating nodules may be visualized as heterogeneous high attenuation, but the sensitivity is rather low.[12]
MRI has greater sensitivity for detection of discrete regenerative nodules than does CT or ultrasound (US).[13] Regenerative nodules usually are isointense with other background nodules on both T1-weighted and T2-weighted images. Less commonly, they may be hyperintense on T1-weighted images and hypointense on T2-weighted images (Figs. 1 and 2). Siderotic regenerating nodules have characteristic imaging features including decreased signal intensity on T2-weighted pulse sequences (Fig. [1]), and demonstrate a blooming effect on gradient echo pulse sequences with longer echo times, which are especially sensitive to magnetic field inhomogeneity, and have high detection sensitivity (Fig. [3]).[14] [15]
On hepatic arterial phase of dynamic CT and MRI after an injection of contrast medium, regenerating nodules show the same degree of enhancement as the liver parenchyma (Fig. [2]).
#Intrahepatic Fibrosis
Cirrhotic parenchyma is characterized at sonography by a coarse, heterogeneous echo pattern, increased parenchymal echogenicity, and increased sound attenuation due to intrahepatic fibrosis, although surface nodularity is a more specific sonographic sign of cirrhosis. When fat deposition coexists with fibrosis, the intrahepatic portal vein is hardly observed.
Only a relatively thick fibrous septum can be detected by CT and MRI because of their limited spatial resolution. Thick fibrous septa and confluent hepatic fibrosis are visualized as low attenuation in plain CT and as low intensity on T1-weighted images (Figs. 1 and 2). However, they are not shown in good contrast to the surrounding liver parenchyma in contrast-enhanced CT because the contrast medium is trapped in the interstitial spaces of hepatic fibrous tissue.[16] They may be shown as high intensity on T2-weighted spin-echo images in MRI because there is invasion by inflammatory cells of the fibrous septum and the presence of pseudo bile ducts.[15]
#Portal Hypertension
The most important clinical manifestations of cirrhosis are related to portal hypertension and include ascites, bleeding from esophageal varices, and hepatic encephalopathy.[1]
Dilatation of the main portal vein and the splenic and superior mesenteric vein can also be seen in some cases. The hepatic artery frequently is enlarged and tortuous in advanced cirrhosis and may demonstrate increased flow.[17] Portosytemic shunt vessels, such as patent paraumbilical veins, dilated coronary veins, gastroesophageal varices, short gastric varices, retroperitoneal and subcutaneous collateral vessels, and spontaneous splenorenal shunts, appear in cirrhosis.
Hepatofugal portal venous flow can also occur in patients with portal hypertension. With progression of the disease, there is subsequent involvement of the central right and left branches and main portal vein itself.[18] In cirrhosis, the phasic oscillations in hepatic venous flow are dampened, presumably due to venous segments that are narrowed by adjacent regenerative nodules or hepatic fibrosis and loss of tissue compliance.[19]
Ultrasonography enhanced by duplex Doppler and color flow Doppler imaging is useful for the evaluation of varices. Duplex sonography is useful for measuring flow velocity and characterizing flow patterns within the hepatic vessels. The varices are also recognized as enhanced tubular structure in contrast-enhanced CT and MRI with a variety of flow-sensitive gradient echo techniques.[20] [21] The technique is usually adequate for assessing patency of the portal and hepatic veins and detecting venous collateral vessels. Furthermore, we can observe three-dimensional structures by CT or MR angiography, using, for example, a maximum intensity projection method or surface rendering method (Fig. [4]).
Gamma-Gandy nodules, which result from bleeding of follicles in portal hypertension, may develop in the spleen; they are visualized as low-intensity nodules in gradient echo images because they contain hemosiderin.[22]
Barium-contrast examinations can show thickened folds in the distal esophagus and fundus of the stomach. Angiography, although invasive, is the primary imaging modality by which the hemodynamic changes in hepatic blood flow in cirrhosis can be measured accurately and is often indicated for the consideration of surgical or nonsurgical intervention.
#IMAGING FEATURES OF NODULAR LESIONS IN CIRRHOSIS
#Hepatocellular Carcinoma (HCC)
#PATHOLOGY OF HCC
Nodular lesions commonly found in cirrhotic livers include regenerative nodules, dysplastic nodules, and hepatocellular carcinoma. Although there is substantial evidence of a temporal progression in cirrhotic livers from regenerative nodule to dysplastic nodule to well-differentiated HCC, carcinoma has also been shown todevelop independently of regenerative or dysplastic nodules.[23]
Eggel's macroscopic classification of HCC is widely known; it shows three types: nodular, massive, and diffuse. In General Rules for the Clinical and Pathological Study of Primary Liver Cancer, issued by the Liver Cancer Study Group of Japan, the nodular type is further divided into three subtypes: single nodule, single nodule with extranodular growth, and contiguous multinodular type.[24] The single nodular type is defined as a carcinoma with a clearly demarcated single nodule, the single nodule with extranodular growth as a single nodular type as a carcinoma with various degrees of infiltrative growth, and the contiguous multinodular type as a nodule formed by the aggregation of several small carcinoma nodules. The massive type is classified as an unclearly and irregularly demarcated carcinoma, and the diffuse type is classified as a carcinoma with thousands of carcinoma nodules all through the liver.[24] Fibrous capsules and septum are often observed in the nodular type of HCC, as are other macroscopic features. Advanced HCC often invades the portal vein (tumor thrombus), and sometimes invades the hepatic vein or bile duct. The imaging findings in HCC reflect these macroscopic features and are very important for diagnosis.
# #EPIDEMIOLOGY OF HCC
Most cases of HCC in Japan are complicated by liver injury, especially cirrhosis. Similarly, a high prevalence of cirrhosis was found in HCC patients elsewhere, for example, Los Angeles (79%), other areas of the United States (74%),[25] Spain (93%),[26] the United Kingdom (75%),[27] and Italy (88%).[28] Because recent advances in medical care have decreased the mortality rate due to cirrhosis itself, a growing fraction of cirrhotic patients may die of HCC rather than from hepatic failure or gastrointestinal bleeding.
The incidence of HCC is rising in the United States and has almost doubled over the past 20 years.[29] This rise is caused in part by the epidemic of hepatitis C virus, which can lead to both cirrhosis and HCC. Cirrhosis due to hepatitis C causes 70% of cases of HCC in Japan and 30 to 50% in the United States,[4] although 40% of the HCC seen in North America occurs in noncirrhotic livers.[30] In patients with hepatitis B virus, HCC can appear in the stage of chronic hepatitis, but the rate did not increase during the age of disease. In patients with hepatitis C virus, the incidence increases in the advanced stage of cirrhosis.
The association between ethanol and HCC has been investigated in thousands of patients. The effect of alcohol is dose-dependent, and cirrhosis could be the basis for ethanol-associated HCC cases. In areas such as the United States, in which the prevalence of hepatitis B virus is low, HCC risk is increases up to 40% with heavy alcohol consumption.[31] In several European countries and Japan, alcohol has been associated with a higher incidence of HCC among hepatitis C patients.[32]
# #IMAGING FEATURES OF HCC
Because the detectability of hepatic nodular lesions on each imaging modality depends on the contrast difference between the normal parenchyma and nodular lesions, the presence of cellularity, fibrosis, fatty change, necrosis, peliosis, and vascularity of nodules can affect the sensitivity of detection and characterization. CT and MRI have been used for the characterization of hepatic tumors detected by US. Ultrasound-guided biopsy to permit histologic examination is sometimes necessary to reach a diagnosis.
The typical findings of ultrasonography of HCC are mosaic pattern, septum formation, peripheral sonolucency (halo), lateral shadow produced by fibrotic pseudocapsule, and posterior echo enhancement (Fig. [5]).[33] When the nodule is small, the internal echo pattern tends to be hypoechoic and uniform (Fig. [6]),[34] whereas large masses are usually heterogeneous (Fig. [5]). Sometimes small, well-differentiated HCC presents with a hyperechoic pattern, which is indicative of fattymetamorphosis, clear cell change, pseudoglandular arrangement of the cancer cells, periodic changes of the vascular space, or sclerotic change in the tumor. In more advanced cases of HCC, portal tumor thrombi, biliary invasion, and hepatic vein invasion are also observed, which strongly indicate the diagnosis of HCC. The detection of multiple nodules is another common finding in HCC.
With the use of the color Doppler ultrasound, real-time imaging combined with blood flow imaging, which is depicted in color, is obtained. A basket pattern in color Doppler images has been described; this pattern represents a fine network of vessels surrounding the tumor nodules (Fig. [5]).[35] The typical color Doppler findings in advanced HCC are afferent pulsatile waveform signals (Fig. [5]), intratumoral pulsatile waveform signals associated with intratumoral continuous waveform signals, and efferent continuous waveform signals.[36] In contrast, the typical findings of early HCC areafferent continuous waveform signals, which reflect a feeding portal flow, rarely associated with pulsatile wave-form signals. Of the several parameters that can be obtained with Doppler spectral analysis, maximum flow velocity (Vmax) and pulsatility index (PI) are very important in the differential diagnosis of hepatic tumors.[37] The PI in HCC is much higher than that in hemangioma.
Power Doppler US is a new technique for depicting a flow signal based on the blood flow, in contrast to the conventional color Doppler, which depicts the flow velocity. Therefore, power Doppler US is very sensitive in the depiction of intratumoral color signal regardless of the velocity or the direction of the blood flow.
A galactose-based US contrast agent for intravenous injection was developed recently[38] and has been applied to various clinical fields including the Doppler study of hepatic tumors. The technique improves the sensitivity of detecting vascularity within the nodule and the efficiency of the differential diagnosis by color Doppler imaging.[39]
Although ultrasonography has been widely used for the diagnosis of HCC, ultrasound examination of the entire liver is occasionally difficult because of intervening bones, air in the intestine or lung, and the presence of fatty tissue. These do not affect CT of the liver, and an advantage of CT over ultrasonography is its objectivity in visualizing the liver. CT has allowed successful detection of large hepatocellular carcinomas (more than 2 cm in diameter) and the differentiation of HCC from other space-occupying lesions.
Most examples of HCC have lower attenuation than the adjacent liver parenchyma on unenhanced CT scans (Fig. [7]). However, small, well-differentiated tumors are often isoattenuating.[40] HCC often has daughter nodules. It can appear on a CT scan as a solitary mass, a dominant mass with daughter lesions, or a diffusely infiltrating neoplasm. It may be multifocal.
To make a definitive diagnosis, we should perform a dynamic study of the whole liver by CT, observing serial images of tumor hemodynamics using contrast media.[41] It is necessary to use a system that can make a helical scan in CT, or make about 15 slices at once by gradient echo imaging in MRI.
In the dynamic study using helical CT, the whole liver in the arterial phase is first imaged when the contrast medium comes from the artery to the liver 25 to 30 sec after the start of injection at a rate of 3 to 5 mL/sec. Sixty to 70 sec after starting, the portal venous phase is imaged when the contrast medium distributed to the bowel is collecting in the liver, and almost 3 min later the equilibrium phase is imaged.[42] However, because of the wide range in circulation time that results from hemodynamic alternations intrinsic to the disease, recent developments using test bolus to determine an individual's circulation time[43] or fluoroscopic ``real time'' triggering have been used to obtain arterial phase imaging more reliably.
HCC-specific hemodynamics are shown to be an increase in arterial blood flow supply and the development of tumor vessels. Small HCC examples usually show diffuse, homogeneous enhancement in the arterial phase, with rapid washout in the portal venous phase (Fig. [8]A).[44] Large examples tend to display a heterogeneous or mosaic pattern of enhancement in the arterial phase and the presence of a capsule and intratumoral septa in the later phases (Fig. [7]).
Arterial-phase CT proved to be the most useful in the detection of hypervascular tumors such as HCC. However, small carcinomas less than 20 mm in diameter in cirrhosis are often well differentiated and are hypovascular relative to the surrounding liver parenchyma. CT images obtained during the portal venous and equilibrium phases have proven to be useful in the detection of less vascularized tumors such as well-differentiated or early HCC.
We have to be careful to differentiate false-positives. A heterogeneous pattern should be distinguished from the transient hepatic attenuation difference (THAD), which is often identified as a focal peripheral wedge of increased intensity seen only on hepatic arterial imaging.[45] These lesions usually are seen in subcapsular areas and do not cause bulging of the capsule. They are usually due to arterioportal shunts or aberrant venous drainage.[46] Arterioportal shunting is often seen in cirrhotic livers without tumor (Fig. [9]).
When a portal tumor thrombus appears as vascular invasion in HCC, the peripheral liver parenchyma shows segmental enhancement because it is compensationally sustained by the hepatic artery.[47] It is possible to differentiate portal tumor thrombus from portal bland thrombus, because the tumor thrombus can be enhanced in the arterial phase. Portal venous-phase scans are also required for assessment of portal vein patency and evaluation of extrahepatic abnormal organs. It is therefore necessary to perform at least adouble-phase dynamic study for the detection and differential diagnosis of HCC.
Regarding signal intensity of classical HCC in MRI, 35% of cases show high intensity on T1-weighted images, 25% show isointensity, and 40% show low intensity, whereas almost all cases other than tumor with coagulation necrosis show high intensity on T2-weighted images (Fig. [10]).[48] Relatively well-differentiated HCC is reported to show high intensity on T1-weighted spin-echo images. One of the reasons is thatfat deposition often appears in well-differentiated HCC. It is also reported that well-differentiated HCC often shows isointensity in T2-weighted spin-echo images.[48] The capsule shows low attenuation on CT, low intensity on T1-weighted spin-echo images (Fig. [10]), and high intensity on T2-weighted spin-echo images, but it may be visualized as a double layer because the outer layer of the capsule containing vessels and microbile ducts distorted by the tumor is imaged at higher intensity than the inner layer composed of fibrous tissue. The intratumoral fibrous septum with less water content is shown at low intensity on T2-weighted images, and the whole tumor is imaged as a heterogeneous high-intensity region. When a tumor is complicated with tumor thrombus, the peripheral liver segment, which may be edematous, shows high intensity on T2-weighted spin-echo images in MRI.[49]
In MRI, we should also perform the dynamic study with T1-weighted gradient echo images during rapid bolus injection of contrast, as with CT. The timing of imaging the dynamic study is almost the same as in CT: The arterial phase is first imaged 25 sec after the start of injection of the contrast medium (Gd-DTPA), and the portal venous and equilibrium phases are continuously imaged. However, there is no significant difference between dynamic helical CT and dynamic MRI in the detection of hypervascular HCC (Fig. [8]B).[41]
MRI evaluation of the cirrhotic liver may be enhanced by the use of targeted MR contrast agents.[50] [51] These agents differ from the extracellular gadolinium chelates in that they are incorporated into hepatic cells or phagocytized by Kupffer cells, and enhance the contrast between the liver parenchyma and the tumor without normal liver cell function or Kupffer cell. Superparamagnetic iron oxide, one of the agents, is useful toevaluate histological tumor grade[52] because well-differentiated HCC often has Kupffer cells, but not poorly differentiated HCC.
#Dysplastic Nodules
#PATHOLOGY OF DYSPLASTIC NODULES
Dysplastic foci are particularly common in patients with cirrhosis secondly to chronic viral hepatitis B and C.[53] Dysplastic nodule refers to a nodular region of hepatocytes 1 mm or greater in diameter with dysplasia but without definite histologic features of malignancy.
Dysplastic nodules are divided into low- and high-grade types. Histologically, low-grade dysplastic nodules contain hepatocytes that are minimally abnormal, without architectural or cytologic atypia, but may contain large cell dysplasia. They usually contain portal tracts and hepatic venules and may retain iron or copper.[54] A high-grade dysplastic nodule may have focal or diffuse architectural or cytologic atypia. A high-grade dysplastic nodule also may contain a microscopic area of HCC (a dysplastic nodule with subfocus of HCC).[55]
Sakamoto et al[56] proposed the development of HCC from a regenerating nodule to a low-grade dysplastic nodule, then to a high-grade dysplastic nodule, and subsequently into well-differentiated HCC and overt HCC in a multistep fashion or in a continuous tradition. Therefore, dysplastic nodules have been considered premalignant lesions.
# #IMAGING FEATURES OF DYSPLASTIC NODULES
There is considerable overlap in the imaging appearances of prominent regenerative nodules, dysplastic nodules, and early HCC.[54]
Dysplastic nodules are shown as homogeneous hypoechoic nodules on ultrasonography. On Doppler ultrasonography, the typical findings of dysplastic nodules are afferent continuous waveform signals, which reflect a feeding portal flow, rarely associated with pulsatile waveform signals.[36]
Some dysplastic nodules are difficult to visualize on dynamic contrast-enhanced images because the blood supply to dysplastic nodules is usually from the portal venous system and they enhance in a manner similar to the surrounding parenchyma. Therefore, marked enhancement during the arterial phase can help distinguish HCC from dysplastic nodules (Fig. [11]).[57] However, the enhancement patterns are less specific for well-differentiated examples of HCC, which often show substantial portal venous enhancement. Moreover, as the grade of malignancy increases, the hepatic arterial flow to nodular lesions tends to increase and the portal venous supply tends to decrease.[58]
The equilibrium-phase contrast-enhanced study can be helpful for distinguishing HCC from regenerating nodules and dysplastic nodules because some carcinomas will have a discernible capsule at this stage, whereas regenerating nodules and dysplastic nodules should not.[59]
The combination of iso- or hyperintensity on T1-weighted images and hypointensity on T2-weighted images has been reported to be specific for dysplastic nodules (Fig. [11]).[60] Possible explanations of high intensity on T1-weighted images include fatty change, intratumoral copper, and increased zinc content of the surrounding hepatic parenchyma.[61]
Dysplastic nodules with foci of HCC are clearly malignant lesions. Doubling times of less than 3 months for the HCC focus have been reported.[62] The focus may show pulsatile waveform signal in Doppler ultrasonography (Fig. [6]) and demonstrates enhancement in the arterial-phase images of dynamic CT and MRI and high intensity on T2-weighted MR images.
#ROLE OF SONOGRAPHIC AND CT SCREENING OF CIRRHOTIC PATIENTS
#Significance
The rate of development of HCC per year in cirrhotic patients was reported as 5.3 to 8.8%.[63] In patients with small HCC (2 cm or less in diameter), the surgical prognosis is significantly better than that in patients with larger nodules. One large study reported that the patient survival rates at 5 yr were 60.5% for nodules 2 cm or less, 39.3% for those 2 to 5cm, and 26.8% for those more than 5 cm.[64] As transplantation techniques have improved, exclusion criteria have become narrow, and patients with early-stage HCC may now undergo transplantation. Thus there is growing interest in the use of screening strategies for the detection of HCC in patients with cirrhotic liver[5] [6] at a relative early, asymptomatic stage, when the disease may respond more favorably to treatment.
#Screening Methods
As mentioned above, modern real-time ultrasound equipment is known to be highly sensitive in the detection of nodular lesions in the liver as small as 0.5 cm in diameter, as well as highly cost effective and easy to perform.[65] Therefore, US has been the method of first choice for screening high-risk patient populations, such as cirrhotic patients and patients infected with hepatitis B or C viruses.
The doubling time of lesions less than 10 mm in diameter is 7 to 8 months, that of lesions 10 to 20 mm 2.2 months,[66] so to detect HCC less than 2 cm, imaging screening should be repeated every 3 months. The doubling time of alpha-fetoprotein concentrations was less than 60 days.[67] In our program, if patients have cirrhosis, they undergo serum alpha-fetoprotein testing and/ or protein induced by vitamin K absence or antagonist II (PIVKA-II) measurements every 2 months, US every 3 months, and CT or MRI every 6 months. Because a small number of patients with chronic hepatitis without cirrhosis develop HCC, these patients are also enrolled in the program and undergo alpha-fetoprotein testing every 4 months, US every 6 months, and CT or MRI every 12 months. This has resulted in great success in detecting small HCC nodules less than 2 cm in diameter and early-stage, well-differentiated HCC, the prognosis in both of which is very good. Based on this screening protocol, at present in Japan approximately 20to 30% of the HCC nodules detected are less than 2cm in diameter, and 50 to 60% are less than 5 cm in diameter.
The alpha-fetoprotein level increased to more than 20 ng/mL in 15 to 20% of the small HCC cases (less than 2 cm), and the positive rate of alpha-fetoprotein was low in well-differentiated HCC cases.[16] Thus regular screening of high-risk patients by US is undoubtedly important in the early detection of small HCC nodules in the clinical setting, especially in countries where HCC is prevalent. Regular follow-up of HCC high-risk patients with US is strongly recommended.
#SUMMARY
Because recent advances in medical care have decreased the mortality rate due to liver cirrhosis itself, many cirrhotic patients die due to HCC. It is therefore very important to know the specific imaging appearances of liver cirrhosis and nodular lesions at the time of imaging evaluation of cirrhotic liver, because HCC in patients at a relatively early asymptomatic stage may respond more favorably to treatment.
#ABBREVIATIONS
CT computed tomography
HCC hepatocellular carcinoma
MRI magnetic resonance imaging
PI pulsatility index
PIVKA-II protein induced by vitamin K absence or antagonist II
THAD transient hepatic attenuation difference
US ultrasound
Vmax maximum flow velocity

Figure 1 Regenerating nodules with and without iron deposition in cirrhotic liver (MRI). Left: T1-weighted gradient echo image. Right: T2-weighted fast spin-echo image. The nodular contour of the liver, as in atrophy of the right lobe and hypertrophy of the left and caudate lobes of the liver, is seen. Regenerative nodules through the liver show hyperintensity on T1-weighted images, and iso- or hypointensity on T2-weighted images. Siderotic regenerating nodules have characteristic imaging features, including decreased signal intensity on T2-weighted image (arrows). Thick fibrous septa are visualized as low intensity on T1-weighted images.

Figure 2 Regenerating nodule in the medial segment. Left: T1-weighted gradient echo MRI. Right: Arterial-phase CT image. The regenerative nodule, showing hyperintensity on T1-weighted image (arrow), reveals the same degree of enhancement as the liver parenchyma in the hepatic arterial phase of dynamic CT.

Figure 3 Siderotic regenerating nodules (MRI). Multiple siderotic regenerating nodules show decreased signal intensity on gradient echo pulse sequence, which is especially sensitive to magnetic field inhomogeneity.

Figure 4 Varices due to portal hypertension. Left: Enhanced CT. Right: CT portography reconstructed from portal venous-phase CT images by using maximum intensity projection method. Dilated coronary veins and gastroesophageal varices are also recognized as enhanced tubular structures in contrast-enhanced CT and as three-dimensional structures by CT angiography.

Figure 5 Poorly differentiated HCC (40 mm in diameter)(US). Left: B-mode image. Middle: Power Doppler image. Right: Pulse Doppler wave. The tumor shows mosaic pattern and peripheral sonolucency (halo) on B-mode image. A basket pattern, representing a fine network of vessels surrounding the tumor nodules (arrows), is seen in color Doppler image. Pulsatile waveform signals, typical color Doppler findings in advanced HCC, are seen within the vessel.

Figure 6 Well-differentiated HCC with an advanced focus (20 mm in diameter). A. Left: B-mode image. Middle: Power Doppler mode image. Right: Pulse Doppler wave. The internal echo of the small nodule is hypoechoic and uniform (arrowheads). On Power Doppler mode, an advanced focus, showing pulsatile waveform signal in pulse Doppler wave, is seen (arrow).

Figure 7 Moderately differentiated HCC. Left: Precontrast image. Middle: Arterial-phase CT image. Right: Equilibrium-phase CT image. The tumor shows lower attenuation than the adjacent liver parenchyma on precontrast CT image. It shows a heterogeneous or mosaic pattern of enhancement on the arterial-phase image and the presence of a capsule on the equilibrium-phase image.

Figure 8 Hypervascular HCC (15 mm in diameter). A. Left: Arterial-phase CT image. Right: Portal venous phase CT image. The tumor shows homogeneous marked enhancement in the arterial phase, with rapid washout in the portal venous phase (arrow), making it invisible in the later images. B. Left: Arterial-phase MRI image. Right: Portal venous phase MRI image. Dynamic MRI shows almost the same enhancement pattern of the tumor as dynamic CT (arrow).

Figure 9 Pseudolesion in cirrhotic liver due to arterioportal shunt. A focal peripheral wedge of enhancement without bulging of the capsule, indicating arterioportal shunt, is seen on the hepatic arterial-phase CT image (arrow). Early portal venous drainage is also seen (arrowheads).

Figure 10 Large HCC. Left: T1-weighted spin-echo imag., Right: T2-weighted spin-echo image. The tumor shows heterogeneous high intensity with low intensity of capsule (arrow), intratumoral septa, and central necrosis on T1-weighted image and heterogeneous high intensity on T2-weighted image.

Figure 11 Dysplastic nodule. Upper left: Arterial-phase CT image. Upper right: Equilibrium-phase CT image. Lower left: T1-weighted gradient echo image. Lower right: T2-weighted fast spin-echo image. The nodule (arrow) does not show marked enhancement on the arterial-phase CT image contrary to HCC. It enhances in a manner similar to the surrounding parenchyma. Capsule is not seen on the equilibrium phase. The nodule (arrowhead) shows hyperintensity on T1-weighted images and isointensity on T2-weighted images.
REFERENCES
- 1 Podolsky D K, Isselbacher K J. Alcohol-related liver disease and cirrhosis. In: Isselbacher KJ, Braunwald E, Wilson JD, et al., eds. Harrison's Principles of Internal Medicine, 13th ed New York: McGraw-Hill, 1994: 1483-1495
- 2 Conn H O, Atterbury C E. Cirrhosis. In: Schiff L, Schiff ER, eds. Diseases of the Liver, 7th ed Philadelphia: Lippincott; 1993 ;875-934
- 3 Genesca J, Esteban J I, Alter H J. Blood-borne non-A, non-B hepatitis: hepatitis C. Semin Liver Dis . 1991; 11 147-164
- 4 Ince N, Wands J R. The increasing incidence of hepatocellular carcinoma. N Engl J Med . 1999; 340 798-799
- 5 Kobayashi K, Sugimoto T, Makino H. Screening methods for early detection of hepatocellular carcinoma. Hepatology . 1985; 5 1100-1105
- 6 Oka H, Kurioka N, Kim K. Prospective study of early detection of hepatocellular carcinoma in patients with cirrhosis. Hepatology . 1990; 12 680-687
- 7 Popper H. Pathologic aspects of cirrhosis: a review. Am J Pathol . 1977; 87 228-264
- 8 Brown J J, Naylor M J, Yagan N. Imaging of hepatic cirrhosis. Radiology . 1997; 202 1-16
- 9 Laennec R TH. Traite de l'auscultation medicale. Paris, France:Chaude, 1826: 196
- 10 Harbin W P, Robert N J, Ferrucci Jr T J. Diagnosis of cirrhosis based on regional changes in hepatic morphology: a radiological and pathological analysis. Radiology . 1980; 135 273-283
- 11 Freeman M P, Vick C W, Taylor K JW. Regenerating nodules in cirrhosis: sonographic appearance with anatomic correlation. AJR . 1986; 146 533-536
- 12 Murakami T, Nakamura H, Hori S. CT and MRI of siderotic regenerating nodules in cirrhotic liver. J Comput Assist Tomogr . 1992; 16 578-582
- 13 Itai Y, Ohnishi S, Ohtomo K. Regenerating nodules of liver cirrhosis: MR imaging with pathologic correlation. AJR . 1987; 165 419-423
- 14 Ohtomo K, Itai Y, Ohtomo Y, Shiga J, Ito M. Regeneration nodules of liver cirrhosis: MR imaging with pathologic correlation. AJR . 1990; 154 505-507
- 15 Murakami T, Kuroda C, Marukawa T. Regenerating nodules of hepatic cirrhosis: MR findings with pathologic correlation. AJR . 1990; 155 1227-1231
- 16 Ohtomo K, Baron R L, Dodd III D G. Confluent hepatic fibrosis in advanced cirrhosis: appearance at CT. Radiology . 1993; 188 31-35
- 17 Ralls P W. Color Doppler sonography of the hepatic artery and portal venous system. AJR . 1990; 155 522-526
- 18 Middleton W D. The liver. In: Kurtz AB, Middleton, eds. Ultrasound: the requisites St. Louis: Mosby-Year Book, 1996: 3-34
- 19 Bolondi L, Li Bassi S, Gaiani S. Liver cirrhosis: changes of Doppler waveform of hepatic veins. Radiology . 1991; 178 513-516
- 20 Edelman R R, Zhao B, Liu C. MR angiography and dynamic flow evaluation of the portal venous system. AJR . 1989; 153 755-760
- 21 Finn J P, Kane R A, Edelman R R. Imaging of the portal venous system in patients with cirrhosis: MR angiography versus duplex Doppler sonography. AJR . 1993; 161 989-994
- 22 Minami M, Itai Y, Ohtomo K. Siderotic nodules in the spleen: MR imaging of portal hypertension. Radiology . 1989; 172 681-684
- 23 Kondo F, Ebara M, Sugiura N. Histological features and clinical course of large regenerative nodules: evaluation of their precancerous potentiality. Hepatology . 1990; 12 592-598
- 24 Liver cancer study group of Japan. General rules for the clinical and pathological study of primary liver cancer
Tokyo: Kanehara, 1992: 14-17
- 25 Craig J R, Klatt E C, Yu M. Role of cirrhosis and the development of HCC: evidence from histologic studies and large population studies. In: Tabor E, Di Bisceglie AM, Purcell RH, eds. Etiology, Pathology and Treatment of Hepatocellular Carcinoma in North America The Woodlands, TX: Portfolio 1991: 177-190
- 26 Calvet X, Bruix J, Bru C. Natural history of hepatocellular carcinoma in Spain. J Hepatol . 1990; 10 311-317
- 27 Zaman S N, Johnson P J, Williams R. Silent cirrhosis in patients with hepatocellular carcinoma: implications for screening in high-incidence and low-incidence areas. Cancer . 1990; 65 1607-1610
- 28 Tribelli C, Melato M, Croce L S. Prevalence of hepatocellular carcinoma and relation to cirrhosis: comparison of two different cities of the world-Trieste, Italy, and Chiba, Japan. Hepatology . 1989; 10 998-1002
- 29 El-Serag H B, Mason A C. Rising incidence of hepatocellular carcinoma in the United States. N Engl J Med . 1999; 340 745-750
- 30 Nzeako V C, Goodman Z D, Ishak K G. Hepatocellular carcinoma in cirrhotic and noncirrhotic livers: a clinico- histopatholgoic study of 804 North American patients. Am J Clin Pathol . 1996; 105 65-75
- 31 Okuda K. Hepatocellular carcinoma in cirrhotics. Semin Liver Dis . 1993; 13 374-383
- 32 Alpert E, Hershberg R, Schur P H. Alpha fetoprotein in human hepatoma: improved detection in serum and quantitative studies using a new sensitive technique. Gastroenterology . 1971; 61 137-143
- 33 Tanaka S, Kitamura T, Imaoka S. Hepatocellular carcinoma: sonographic and histologic correlation. AJR . 1983; 140 701-707
- 34 Dodd III D G, Miller W J, Baron R L, Kolnick M L, Campbell W L. Detection of malignant tumors in end-stage cirrhotic livers: efficacy of sonography as a screening tool. AJR . 1992; 159 727-733
- 35 Tanaka S, Kitamura T, Fujita M, Nakanishi K, Okuda S. Color Doppler flow imaging of liver tumors. AJR . 1990; 154 509-514
- 36 Tanaka S, Kitamura T, Fujita M. Small hepatocellular carcinoma: differentiation from adenomatous hyperplastic nodule with color Doppler flow imaging. Radiology . 1992; 182 161-165
- 37 Tochio H, Minowa K, Tomita S. Differential diagnosis of hepatic tumors using color Doppler flow imaging: value of the Doppler spectral analysis. Jpn J Med Ultrasonics . 1992; 19 277-287
- 38 Nanda N C, Schlief R. Advances in Echo Imaging Using Contrast Enhancement. Dordrecht, the Netherlands: Kluwer Academic Publishers; 1993
- 39 Fujimoto M, Moriyasu F, Nishikawa K. Color Doppler sonography of hepatic tumors with a galactose-based contrast agent: Correlation with angiographic findings. AJR . 1994; 163 1099-1104
- 40 Yamashita Y, Takahashi M, Baba Y. Hepatocellular carcinoma with or without cirrhosis: a comparison of CT and angiographic presentations in the United States and Japan. Abdom Imaging . 1993; 18 168-175
- 41 Kim T, Murakami T, Oi H. Detection of hypervascular hepatocellular carcinoma by dynamic MRI and dynamic spiral CT. J Comput Assist Tomogr . 1995; 19 948-954
- 42 Mitsuzaki K, Yamashita Y, Ogata I, Nishiharu T, Urata J, Takahashi M. Multiple-phase helical CT of the liver for detecting small hepatomas in patients with liver cirrhosis: contrast-injection protocol and optimal timing. AJR . 1996; 167 753-757
- 43 Earls J, Rofsky N R, DeCorato D. Arterial-phase dynamic gadolinium-enhanced MR imaging: optimizing with a test examination and a power injector. Radiology . 1997; 202 268-273
- 44 Kelekis N L, Semelka R C, Worawattanakul S. Hepatocellular carcinoma in North America: a multiinstitutional study of appearance on T1-weighted, T2-weighted, and serial gadolinium-enhanced gradient-echo images. AJR . 1998; 170 1005-1013
- 45 Itai Y, Matsui O. Blood flow and liver imaging. Radiology . 1997; 202 306-314
- 46 Matsui O, Kadoya M, Yoshikawa J. Aberrant gastric venous drainage in cirrhotic livers: imaging findings in focal areas of liver parenchyma. Radiology . 1995; 197 345-349
- 47 Mitani T, Nakamura H, Murakami T. Dynamic MR studies of hepatocellular carcinoma with portal vein tumor thrombosis. Radiation Medicine . 1992; 10 232-234
- 48 Kadoya M, Matsui O, Takashima T, Nonomura A. Hepatocellular carcinoma: correlation of MR imaging and histologic findings. Radiology . 1992; 183 819-825
- 49 Itai Y, Ohtomo K, Furui S, Minami M, Yoshikawa K, Yashiro N. Lobar intensity differences of the liver on MR imaging. J Comput Assist Tomogr . 1986; 10 236-241
- 50 Weissleder R. Liver MR imaging with iron oxides: toward consensus and clinical practice. Radiology . 1994; 193 593-595
- 51 Murakami T, Baron R L, Federle M P. Cirrhosis of the liver: MR imaging with mangafodipir trisodium (Mn-DPDP). Radiology . 1996; 200 69-77
- 52 Imai Y, Murakami T, Yoshida S. Superparamagnetic iron oxide-enhanced MR images of hepatocellular carcinoma: correlation with histological grading. Hepatology . 2000; 32 205-212
- 53 International working party. Terminology of nodular hepatocellular lesions. Hepatology . 1995; 22 983-993
- 54 Hytiroglou P, Theis N D. Differential diagnosis of hepatocellular nodular lesions. Semin Diagn Pathol . 1998; 15 285-299
- 55 Mitchell D G. Focal manifestations of diffuse liver disease at MR imaging. Radiology . 1992; 185 1-11
- 56 Sakamoto M, Hirohashi S, Shimosato Y. Early stages of multistep hepatocarcinogenesis: adenomatous hyperplasia and early hepatocellular carcinoma. Hum Pathol . 1991; 22 172-178
- 57 Lim J H, Kim C K, Lee W J. Detection of hepatocellular carcinoma and dysplastic nodules in cirrhotic livers: accuracy of helical CT in transplant patients. AJR . 2000; 175 693-698
- 58 Matsui O, Kadoya M, Kameyama T. Benign and malignant nodules in cirrhotic livers: distinction based on blood supply. Radiology . 1991; 178 493-497
- 59 Lim J H, Kim E Y, Lee W J. Regenerating nodules in liver cirrhosis: findings at CT during arterial portography and CT hepatic arteriography with histopathologic correlation. Radiology . 1999; 210 451-458
- 60 Ebara M, Ohto M, Waranabe Y. Diagnosis of small hepatocellular carcinoma: correlation of MR imaging and tumor histologic studies. Radiology . 1986; 159 371-377
- 61 Ebara M, Fukuda H, Kojima Y. Small hepatocellular carcinoma: relationship of signal intensity to histopathologic findings and metal content of the tumor and surrounding hepatic parenchyma. Radiology . 1999; 210 81-88
- 62 Sedek A G, Mitchell D G, Siegelman E S. Early hepatocellular carcinoma that develops within macroregenerative nodules: growth rate depicted at serial MR imaging. Radiology . 1995; 195 753-756
- 63 Oka H, Kurioka N, Kanno T. Prospective study of early detection of hepatocellular carcinoma in patient with cirrhosis. Hepatology . 1990; 12 680-687
- 64 Arii S, Tobe T. Results of surgical treatment. Follow-up study by liver cancer study group of Japan In: Tobe T, et al, eds. Primary Liver Cancer in Japan Tokyo: Springer-Verlag 1992: 243-255
- 65 Tanaka S, Kitamura T, Nakanishi K. Recent advances in ultrasonographic diagnosis of hepatocellular carcinoma. Cancer . 1989; 63 1313-1317
- 66 Ebara M, Ohto M, Shinagawa T. Natural history of minute hepatocellular carcinoma smaller than three centimeters complicating cirrhosis: a study in 22 patients. Gastroenterology . 1986; 90 289-298
- 67 Kaneko S, Unoura M, Kobayashi K. Early detection of hepatocellular carcinoma. In: Okuda K, Tabor E, eds. Liver Cancer New York: Churchill Livingstone 1997: 393-406
REFERENCES
- 1 Podolsky D K, Isselbacher K J. Alcohol-related liver disease and cirrhosis. In: Isselbacher KJ, Braunwald E, Wilson JD, et al., eds. Harrison's Principles of Internal Medicine, 13th ed New York: McGraw-Hill, 1994: 1483-1495
- 2 Conn H O, Atterbury C E. Cirrhosis. In: Schiff L, Schiff ER, eds. Diseases of the Liver, 7th ed Philadelphia: Lippincott; 1993 ;875-934
- 3 Genesca J, Esteban J I, Alter H J. Blood-borne non-A, non-B hepatitis: hepatitis C. Semin Liver Dis . 1991; 11 147-164
- 4 Ince N, Wands J R. The increasing incidence of hepatocellular carcinoma. N Engl J Med . 1999; 340 798-799
- 5 Kobayashi K, Sugimoto T, Makino H. Screening methods for early detection of hepatocellular carcinoma. Hepatology . 1985; 5 1100-1105
- 6 Oka H, Kurioka N, Kim K. Prospective study of early detection of hepatocellular carcinoma in patients with cirrhosis. Hepatology . 1990; 12 680-687
- 7 Popper H. Pathologic aspects of cirrhosis: a review. Am J Pathol . 1977; 87 228-264
- 8 Brown J J, Naylor M J, Yagan N. Imaging of hepatic cirrhosis. Radiology . 1997; 202 1-16
- 9 Laennec R TH. Traite de l'auscultation medicale. Paris, France:Chaude, 1826: 196
- 10 Harbin W P, Robert N J, Ferrucci Jr T J. Diagnosis of cirrhosis based on regional changes in hepatic morphology: a radiological and pathological analysis. Radiology . 1980; 135 273-283
- 11 Freeman M P, Vick C W, Taylor K JW. Regenerating nodules in cirrhosis: sonographic appearance with anatomic correlation. AJR . 1986; 146 533-536
- 12 Murakami T, Nakamura H, Hori S. CT and MRI of siderotic regenerating nodules in cirrhotic liver. J Comput Assist Tomogr . 1992; 16 578-582
- 13 Itai Y, Ohnishi S, Ohtomo K. Regenerating nodules of liver cirrhosis: MR imaging with pathologic correlation. AJR . 1987; 165 419-423
- 14 Ohtomo K, Itai Y, Ohtomo Y, Shiga J, Ito M. Regeneration nodules of liver cirrhosis: MR imaging with pathologic correlation. AJR . 1990; 154 505-507
- 15 Murakami T, Kuroda C, Marukawa T. Regenerating nodules of hepatic cirrhosis: MR findings with pathologic correlation. AJR . 1990; 155 1227-1231
- 16 Ohtomo K, Baron R L, Dodd III D G. Confluent hepatic fibrosis in advanced cirrhosis: appearance at CT. Radiology . 1993; 188 31-35
- 17 Ralls P W. Color Doppler sonography of the hepatic artery and portal venous system. AJR . 1990; 155 522-526
- 18 Middleton W D. The liver. In: Kurtz AB, Middleton, eds. Ultrasound: the requisites St. Louis: Mosby-Year Book, 1996: 3-34
- 19 Bolondi L, Li Bassi S, Gaiani S. Liver cirrhosis: changes of Doppler waveform of hepatic veins. Radiology . 1991; 178 513-516
- 20 Edelman R R, Zhao B, Liu C. MR angiography and dynamic flow evaluation of the portal venous system. AJR . 1989; 153 755-760
- 21 Finn J P, Kane R A, Edelman R R. Imaging of the portal venous system in patients with cirrhosis: MR angiography versus duplex Doppler sonography. AJR . 1993; 161 989-994
- 22 Minami M, Itai Y, Ohtomo K. Siderotic nodules in the spleen: MR imaging of portal hypertension. Radiology . 1989; 172 681-684
- 23 Kondo F, Ebara M, Sugiura N. Histological features and clinical course of large regenerative nodules: evaluation of their precancerous potentiality. Hepatology . 1990; 12 592-598
- 24 Liver cancer study group of Japan. General rules for the clinical and pathological study of primary liver cancer
Tokyo: Kanehara, 1992: 14-17
- 25 Craig J R, Klatt E C, Yu M. Role of cirrhosis and the development of HCC: evidence from histologic studies and large population studies. In: Tabor E, Di Bisceglie AM, Purcell RH, eds. Etiology, Pathology and Treatment of Hepatocellular Carcinoma in North America The Woodlands, TX: Portfolio 1991: 177-190
- 26 Calvet X, Bruix J, Bru C. Natural history of hepatocellular carcinoma in Spain. J Hepatol . 1990; 10 311-317
- 27 Zaman S N, Johnson P J, Williams R. Silent cirrhosis in patients with hepatocellular carcinoma: implications for screening in high-incidence and low-incidence areas. Cancer . 1990; 65 1607-1610
- 28 Tribelli C, Melato M, Croce L S. Prevalence of hepatocellular carcinoma and relation to cirrhosis: comparison of two different cities of the world-Trieste, Italy, and Chiba, Japan. Hepatology . 1989; 10 998-1002
- 29 El-Serag H B, Mason A C. Rising incidence of hepatocellular carcinoma in the United States. N Engl J Med . 1999; 340 745-750
- 30 Nzeako V C, Goodman Z D, Ishak K G. Hepatocellular carcinoma in cirrhotic and noncirrhotic livers: a clinico- histopatholgoic study of 804 North American patients. Am J Clin Pathol . 1996; 105 65-75
- 31 Okuda K. Hepatocellular carcinoma in cirrhotics. Semin Liver Dis . 1993; 13 374-383
- 32 Alpert E, Hershberg R, Schur P H. Alpha fetoprotein in human hepatoma: improved detection in serum and quantitative studies using a new sensitive technique. Gastroenterology . 1971; 61 137-143
- 33 Tanaka S, Kitamura T, Imaoka S. Hepatocellular carcinoma: sonographic and histologic correlation. AJR . 1983; 140 701-707
- 34 Dodd III D G, Miller W J, Baron R L, Kolnick M L, Campbell W L. Detection of malignant tumors in end-stage cirrhotic livers: efficacy of sonography as a screening tool. AJR . 1992; 159 727-733
- 35 Tanaka S, Kitamura T, Fujita M, Nakanishi K, Okuda S. Color Doppler flow imaging of liver tumors. AJR . 1990; 154 509-514
- 36 Tanaka S, Kitamura T, Fujita M. Small hepatocellular carcinoma: differentiation from adenomatous hyperplastic nodule with color Doppler flow imaging. Radiology . 1992; 182 161-165
- 37 Tochio H, Minowa K, Tomita S. Differential diagnosis of hepatic tumors using color Doppler flow imaging: value of the Doppler spectral analysis. Jpn J Med Ultrasonics . 1992; 19 277-287
- 38 Nanda N C, Schlief R. Advances in Echo Imaging Using Contrast Enhancement. Dordrecht, the Netherlands: Kluwer Academic Publishers; 1993
- 39 Fujimoto M, Moriyasu F, Nishikawa K. Color Doppler sonography of hepatic tumors with a galactose-based contrast agent: Correlation with angiographic findings. AJR . 1994; 163 1099-1104
- 40 Yamashita Y, Takahashi M, Baba Y. Hepatocellular carcinoma with or without cirrhosis: a comparison of CT and angiographic presentations in the United States and Japan. Abdom Imaging . 1993; 18 168-175
- 41 Kim T, Murakami T, Oi H. Detection of hypervascular hepatocellular carcinoma by dynamic MRI and dynamic spiral CT. J Comput Assist Tomogr . 1995; 19 948-954
- 42 Mitsuzaki K, Yamashita Y, Ogata I, Nishiharu T, Urata J, Takahashi M. Multiple-phase helical CT of the liver for detecting small hepatomas in patients with liver cirrhosis: contrast-injection protocol and optimal timing. AJR . 1996; 167 753-757
- 43 Earls J, Rofsky N R, DeCorato D. Arterial-phase dynamic gadolinium-enhanced MR imaging: optimizing with a test examination and a power injector. Radiology . 1997; 202 268-273
- 44 Kelekis N L, Semelka R C, Worawattanakul S. Hepatocellular carcinoma in North America: a multiinstitutional study of appearance on T1-weighted, T2-weighted, and serial gadolinium-enhanced gradient-echo images. AJR . 1998; 170 1005-1013
- 45 Itai Y, Matsui O. Blood flow and liver imaging. Radiology . 1997; 202 306-314
- 46 Matsui O, Kadoya M, Yoshikawa J. Aberrant gastric venous drainage in cirrhotic livers: imaging findings in focal areas of liver parenchyma. Radiology . 1995; 197 345-349
- 47 Mitani T, Nakamura H, Murakami T. Dynamic MR studies of hepatocellular carcinoma with portal vein tumor thrombosis. Radiation Medicine . 1992; 10 232-234
- 48 Kadoya M, Matsui O, Takashima T, Nonomura A. Hepatocellular carcinoma: correlation of MR imaging and histologic findings. Radiology . 1992; 183 819-825
- 49 Itai Y, Ohtomo K, Furui S, Minami M, Yoshikawa K, Yashiro N. Lobar intensity differences of the liver on MR imaging. J Comput Assist Tomogr . 1986; 10 236-241
- 50 Weissleder R. Liver MR imaging with iron oxides: toward consensus and clinical practice. Radiology . 1994; 193 593-595
- 51 Murakami T, Baron R L, Federle M P. Cirrhosis of the liver: MR imaging with mangafodipir trisodium (Mn-DPDP). Radiology . 1996; 200 69-77
- 52 Imai Y, Murakami T, Yoshida S. Superparamagnetic iron oxide-enhanced MR images of hepatocellular carcinoma: correlation with histological grading. Hepatology . 2000; 32 205-212
- 53 International working party. Terminology of nodular hepatocellular lesions. Hepatology . 1995; 22 983-993
- 54 Hytiroglou P, Theis N D. Differential diagnosis of hepatocellular nodular lesions. Semin Diagn Pathol . 1998; 15 285-299
- 55 Mitchell D G. Focal manifestations of diffuse liver disease at MR imaging. Radiology . 1992; 185 1-11
- 56 Sakamoto M, Hirohashi S, Shimosato Y. Early stages of multistep hepatocarcinogenesis: adenomatous hyperplasia and early hepatocellular carcinoma. Hum Pathol . 1991; 22 172-178
- 57 Lim J H, Kim C K, Lee W J. Detection of hepatocellular carcinoma and dysplastic nodules in cirrhotic livers: accuracy of helical CT in transplant patients. AJR . 2000; 175 693-698
- 58 Matsui O, Kadoya M, Kameyama T. Benign and malignant nodules in cirrhotic livers: distinction based on blood supply. Radiology . 1991; 178 493-497
- 59 Lim J H, Kim E Y, Lee W J. Regenerating nodules in liver cirrhosis: findings at CT during arterial portography and CT hepatic arteriography with histopathologic correlation. Radiology . 1999; 210 451-458
- 60 Ebara M, Ohto M, Waranabe Y. Diagnosis of small hepatocellular carcinoma: correlation of MR imaging and tumor histologic studies. Radiology . 1986; 159 371-377
- 61 Ebara M, Fukuda H, Kojima Y. Small hepatocellular carcinoma: relationship of signal intensity to histopathologic findings and metal content of the tumor and surrounding hepatic parenchyma. Radiology . 1999; 210 81-88
- 62 Sedek A G, Mitchell D G, Siegelman E S. Early hepatocellular carcinoma that develops within macroregenerative nodules: growth rate depicted at serial MR imaging. Radiology . 1995; 195 753-756
- 63 Oka H, Kurioka N, Kanno T. Prospective study of early detection of hepatocellular carcinoma in patient with cirrhosis. Hepatology . 1990; 12 680-687
- 64 Arii S, Tobe T. Results of surgical treatment. Follow-up study by liver cancer study group of Japan In: Tobe T, et al, eds. Primary Liver Cancer in Japan Tokyo: Springer-Verlag 1992: 243-255
- 65 Tanaka S, Kitamura T, Nakanishi K. Recent advances in ultrasonographic diagnosis of hepatocellular carcinoma. Cancer . 1989; 63 1313-1317
- 66 Ebara M, Ohto M, Shinagawa T. Natural history of minute hepatocellular carcinoma smaller than three centimeters complicating cirrhosis: a study in 22 patients. Gastroenterology . 1986; 90 289-298
- 67 Kaneko S, Unoura M, Kobayashi K. Early detection of hepatocellular carcinoma. In: Okuda K, Tabor E, eds. Liver Cancer New York: Churchill Livingstone 1997: 393-406

Figure 1 Regenerating nodules with and without iron deposition in cirrhotic liver (MRI). Left: T1-weighted gradient echo image. Right: T2-weighted fast spin-echo image. The nodular contour of the liver, as in atrophy of the right lobe and hypertrophy of the left and caudate lobes of the liver, is seen. Regenerative nodules through the liver show hyperintensity on T1-weighted images, and iso- or hypointensity on T2-weighted images. Siderotic regenerating nodules have characteristic imaging features, including decreased signal intensity on T2-weighted image (arrows). Thick fibrous septa are visualized as low intensity on T1-weighted images.

Figure 2 Regenerating nodule in the medial segment. Left: T1-weighted gradient echo MRI. Right: Arterial-phase CT image. The regenerative nodule, showing hyperintensity on T1-weighted image (arrow), reveals the same degree of enhancement as the liver parenchyma in the hepatic arterial phase of dynamic CT.

Figure 3 Siderotic regenerating nodules (MRI). Multiple siderotic regenerating nodules show decreased signal intensity on gradient echo pulse sequence, which is especially sensitive to magnetic field inhomogeneity.

Figure 4 Varices due to portal hypertension. Left: Enhanced CT. Right: CT portography reconstructed from portal venous-phase CT images by using maximum intensity projection method. Dilated coronary veins and gastroesophageal varices are also recognized as enhanced tubular structures in contrast-enhanced CT and as three-dimensional structures by CT angiography.

Figure 5 Poorly differentiated HCC (40 mm in diameter)(US). Left: B-mode image. Middle: Power Doppler image. Right: Pulse Doppler wave. The tumor shows mosaic pattern and peripheral sonolucency (halo) on B-mode image. A basket pattern, representing a fine network of vessels surrounding the tumor nodules (arrows), is seen in color Doppler image. Pulsatile waveform signals, typical color Doppler findings in advanced HCC, are seen within the vessel.

Figure 6 Well-differentiated HCC with an advanced focus (20 mm in diameter). A. Left: B-mode image. Middle: Power Doppler mode image. Right: Pulse Doppler wave. The internal echo of the small nodule is hypoechoic and uniform (arrowheads). On Power Doppler mode, an advanced focus, showing pulsatile waveform signal in pulse Doppler wave, is seen (arrow).

Figure 7 Moderately differentiated HCC. Left: Precontrast image. Middle: Arterial-phase CT image. Right: Equilibrium-phase CT image. The tumor shows lower attenuation than the adjacent liver parenchyma on precontrast CT image. It shows a heterogeneous or mosaic pattern of enhancement on the arterial-phase image and the presence of a capsule on the equilibrium-phase image.

Figure 8 Hypervascular HCC (15 mm in diameter). A. Left: Arterial-phase CT image. Right: Portal venous phase CT image. The tumor shows homogeneous marked enhancement in the arterial phase, with rapid washout in the portal venous phase (arrow), making it invisible in the later images. B. Left: Arterial-phase MRI image. Right: Portal venous phase MRI image. Dynamic MRI shows almost the same enhancement pattern of the tumor as dynamic CT (arrow).

Figure 9 Pseudolesion in cirrhotic liver due to arterioportal shunt. A focal peripheral wedge of enhancement without bulging of the capsule, indicating arterioportal shunt, is seen on the hepatic arterial-phase CT image (arrow). Early portal venous drainage is also seen (arrowheads).

Figure 10 Large HCC. Left: T1-weighted spin-echo imag., Right: T2-weighted spin-echo image. The tumor shows heterogeneous high intensity with low intensity of capsule (arrow), intratumoral septa, and central necrosis on T1-weighted image and heterogeneous high intensity on T2-weighted image.

Figure 11 Dysplastic nodule. Upper left: Arterial-phase CT image. Upper right: Equilibrium-phase CT image. Lower left: T1-weighted gradient echo image. Lower right: T2-weighted fast spin-echo image. The nodule (arrow) does not show marked enhancement on the arterial-phase CT image contrary to HCC. It enhances in a manner similar to the surrounding parenchyma. Capsule is not seen on the equilibrium phase. The nodule (arrowhead) shows hyperintensity on T1-weighted images and isointensity on T2-weighted images.