Multiphase contrast-enhanced magnetic resonance imaging (MRI) is the current modality of choice for characterization of liver masses incidentally detected on imaging. Contrast-enhanced computed tomography (CT) performed in the portal phase is the mainstay for the screening of liver metastases. Characterization of a liver mass by CT and MRI primarily relies on the dynamic contrast-enhancement characteristics of the mass in multiple phases. Noninvasive MRI and CT imaging characteristics of benign and malignant liver masses, coupled with relevant clinical information, allow reliable characterization of most liver lesions. Some cases may have nonspecific or overlapping features that may present a diagnostic dilemma.
Key points
- •
Often multiple imaging modalities may be necessary to characterize liver lesions.
- •
Each modality has unique advantages and disadvantages.
- •
A knowledge of the common benign and malignant lesions observed in the liver is necessary for optimal differential diagnosis and subsequently management.
- •
Imaging is crucial along the entire trajectory of the management of patient with malignant lesions in the liver.
- •
A interdisciplinary team is requisite to obtain optimal oncologic outcomes.
Introduction
In patients without a known extrahepatic malignancy, a hepatic mass may be discovered incidentally on ultrasonography, computed tomography (CT) or magnetic resonance imaging (MRI). Metastatic disease should always be considered in the differential for a mass that does not meet imaging criteria for a simple cyst in a patient with known extrahepatic malignancy undergoing imaging surveillance. However, not infrequently these masses can represent an incidental benign mass such as a hemangioma or focal nodular hyperplasia (FNH). In patients with chronic liver disease or cirrhosis, a hepatic mass may be detected during imaging surveillance. Although hepatocellular carcinoma (HCC) is the leading differential for such masses, benign masses can occur in the cirrhotic liver, and nodules less than 2 cm in diameter in the cirrhotic liver frequently represent regenerating or dysplastic nodules.
Imaging, especially MRI and CT enhanced by contrast material, is instrumental in noninvasive characterization of a liver mass. The American College of Radiology (ACR) Appropriateness Criteria are evidence-based guidelines developed by experts in the field to guide referring physicians in choosing the most appropriate imaging test for a specific clinical condition. Advances in MRI now allows rapid imaging and 3-dimensional acquisition, which, coupled with the soft-tissue contrast, renders MRI the imaging standard for noninvasive characterization of focal liver masses, also endorsed by the ACR. The ACR Appropriateness Criteria guidelines for initial characterization of a focal liver lesion larger than 1 cm encountered in different clinical and imaging scenarios are summarized in Table 1 . 18 F-Fluorodeoxyflucose (FDG) positron emission tomography (PET) combined with CT has an ancillary role in the evaluation of liver metastases if the primary tumor is FDG-avid. Conventional catheter angiography and CT hepatic arteriography and portography, used historically, are no longer used for evaluation of liver masses. Technetium-99m ( 99m Tc) sulfur colloid scan and 99m Tc red blood cell (RBC) scintigraphy are rarely used for evaluation of liver masses. The choice of an imaging modality can vary significantly across institutions based on local radiologic expertise, availability of equipment, and the wishes and biases of treating physicians and radiologists.
| Clinical Variants | Lesion Initially Identified on US | Rating | Lesion Initially Identified on CT | Rating | Lesion Initially Identified on MRI Without Contrast | Rating | Lesion Initially Identified on US, CT, or MRI | Rating |
|---|---|---|---|---|---|---|---|---|
| Normal liver. (no suspicion or evidence of extrahepatic malignancy or liver disease) | MRI abdomen without and with contrast (examination of choice) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | Percutaneous biopsy liver (if lesion remains indeterminate after contrast-enhanced CT or MRI) | 5 |
| CT abdomen without and with contrast (if not cystic on US) | 7 | US abdomen (decide cyst vs solid) | 6 | CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | |||
| US abdomen (if CT contrast contraindicated) | 6 | |||||||
| Known history of extrahepatic malignancy | MRI abdomen without and with contrast (if not cystic on US) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | Percutaneous biopsy liver (if lesion remains indeterminate) | 7 |
| CT abdomen without and with contrast | 7 | US abdomen (decide cyst vs solid) | 6 | CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | FDG-PET/CT whole body (for complete staging if FDG-avid primary) | 6 | |
| US abdomen (decide cyst vs solid or confirm hemangioma) | 6 | |||||||
| Known or suspected liver disease associated with a high risk of hepatocellular carcinoma (chronic hepatitis, cirrhosis, hemochromatosis, etc) | MRI abdomen without and with contrast (examination of choice for surveillance in young patient with hepatitis B or C) | 9 | MRI abdomen without and with contrast (if CT not conclusive or following treatments) | 9 | MRI abdomen without and with contrast (if further characterization needed) | 9 | Percutaneous biopsy liver (if features are not typical) | 6 |
| CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | US abdomen (decide cyst vs solid or to evaluate for biopsy or RFA) | 5 | CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | |||
| US abdomen (to confirm cyst if lesion bright on T2) | 5 |
Knowledge of the underlying key pathologic features and imaging findings of liver masses on MRI and CT allows characterization in most cases. Some masses, however, may exhibit overlapping and nonspecific radiologic features, and in such cases percutaneous image-guided biopsy may become necessary. This article discusses the typical gross morphologic and imaging features of malignant liver masses and certain benign liver masses that may mimic malignancy, preceded by a brief overview of the imaging techniques in current use.
Introduction
In patients without a known extrahepatic malignancy, a hepatic mass may be discovered incidentally on ultrasonography, computed tomography (CT) or magnetic resonance imaging (MRI). Metastatic disease should always be considered in the differential for a mass that does not meet imaging criteria for a simple cyst in a patient with known extrahepatic malignancy undergoing imaging surveillance. However, not infrequently these masses can represent an incidental benign mass such as a hemangioma or focal nodular hyperplasia (FNH). In patients with chronic liver disease or cirrhosis, a hepatic mass may be detected during imaging surveillance. Although hepatocellular carcinoma (HCC) is the leading differential for such masses, benign masses can occur in the cirrhotic liver, and nodules less than 2 cm in diameter in the cirrhotic liver frequently represent regenerating or dysplastic nodules.
Imaging, especially MRI and CT enhanced by contrast material, is instrumental in noninvasive characterization of a liver mass. The American College of Radiology (ACR) Appropriateness Criteria are evidence-based guidelines developed by experts in the field to guide referring physicians in choosing the most appropriate imaging test for a specific clinical condition. Advances in MRI now allows rapid imaging and 3-dimensional acquisition, which, coupled with the soft-tissue contrast, renders MRI the imaging standard for noninvasive characterization of focal liver masses, also endorsed by the ACR. The ACR Appropriateness Criteria guidelines for initial characterization of a focal liver lesion larger than 1 cm encountered in different clinical and imaging scenarios are summarized in Table 1 . 18 F-Fluorodeoxyflucose (FDG) positron emission tomography (PET) combined with CT has an ancillary role in the evaluation of liver metastases if the primary tumor is FDG-avid. Conventional catheter angiography and CT hepatic arteriography and portography, used historically, are no longer used for evaluation of liver masses. Technetium-99m ( 99m Tc) sulfur colloid scan and 99m Tc red blood cell (RBC) scintigraphy are rarely used for evaluation of liver masses. The choice of an imaging modality can vary significantly across institutions based on local radiologic expertise, availability of equipment, and the wishes and biases of treating physicians and radiologists.
| Clinical Variants | Lesion Initially Identified on US | Rating | Lesion Initially Identified on CT | Rating | Lesion Initially Identified on MRI Without Contrast | Rating | Lesion Initially Identified on US, CT, or MRI | Rating |
|---|---|---|---|---|---|---|---|---|
| Normal liver. (no suspicion or evidence of extrahepatic malignancy or liver disease) | MRI abdomen without and with contrast (examination of choice) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | Percutaneous biopsy liver (if lesion remains indeterminate after contrast-enhanced CT or MRI) | 5 |
| CT abdomen without and with contrast (if not cystic on US) | 7 | US abdomen (decide cyst vs solid) | 6 | CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | |||
| US abdomen (if CT contrast contraindicated) | 6 | |||||||
| Known history of extrahepatic malignancy | MRI abdomen without and with contrast (if not cystic on US) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | MRI abdomen without and with contrast (if further characterization needed) | 8 | Percutaneous biopsy liver (if lesion remains indeterminate) | 7 |
| CT abdomen without and with contrast | 7 | US abdomen (decide cyst vs solid) | 6 | CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | FDG-PET/CT whole body (for complete staging if FDG-avid primary) | 6 | |
| US abdomen (decide cyst vs solid or confirm hemangioma) | 6 | |||||||
| Known or suspected liver disease associated with a high risk of hepatocellular carcinoma (chronic hepatitis, cirrhosis, hemochromatosis, etc) | MRI abdomen without and with contrast (examination of choice for surveillance in young patient with hepatitis B or C) | 9 | MRI abdomen without and with contrast (if CT not conclusive or following treatments) | 9 | MRI abdomen without and with contrast (if further characterization needed) | 9 | Percutaneous biopsy liver (if features are not typical) | 6 |
| CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | US abdomen (decide cyst vs solid or to evaluate for biopsy or RFA) | 5 | CT abdomen without and with contrast (if gadolinium contraindicated) | 7 | |||
| US abdomen (to confirm cyst if lesion bright on T2) | 5 |
Knowledge of the underlying key pathologic features and imaging findings of liver masses on MRI and CT allows characterization in most cases. Some masses, however, may exhibit overlapping and nonspecific radiologic features, and in such cases percutaneous image-guided biopsy may become necessary. This article discusses the typical gross morphologic and imaging features of malignant liver masses and certain benign liver masses that may mimic malignancy, preceded by a brief overview of the imaging techniques in current use.
Magnetic resonance imaging
MRI has high sensitivity and specificity for both detection and characterization of benign and malignant focal liver masses. An important advantage of MRI over CT is the lack of ionizing radiation. However, disadvantages include greater cost, longer imaging times, and higher frequency of suboptimal imaging caused by motion artifacts, particularly in patients who cannot perform adequate breath-holding (15–20 seconds).
Of the variety of different protocols that exist for imaging the liver with MRI, a group of core pulse sequences are routinely obtained. The first of these is most often a set of T2-weighted images. Fluid is hyperintense on T2-weighted imaging, allowing for identification of cysts and cystic masses. Other lesions such as hemangiomas are typically markedly intense (slightly less so than cysts) on T2-weighted images. Both benign and malignant solid tumors may be mildly to moderately hyperintense, but T2-weighted imaging alone is neither highly sensitive nor specific in characterizing focal liver lesions. All liver protocols should also include T1-weighted “in and out of phase” imaging. These sequences are used to identify tissues with internal microscopic fat; which can be seen in some hepatic masses such as hepatocellular adenomas and HCCs. The mainstay of liver imaging with MRI is dynamic contrast-enhanced fat-saturated T1-weighted imaging using a gadolinium chelate. Conventional extracellular gadolinium-based contrast agents are analogous to iodinated contrast used in CT, and lesions will follow similar enhancement patterns on both modalities. First, precontrast images are acquired, which provide information regarding T1 characteristics of lesions (internal hemorrhage showing increased signal intensity) and serve as a baseline to evaluate for contrast enhancement. Following this, at least 3 dynamic acquisitions are obtained in the arterial, portal venous, and equilibrium phases after intravenous injection of a gadolinium-based contrast agent. A 10- to 20-minute delayed postcontrast image may also be obtained to evaluate a slowly filling hemangioma or for delayed enhancement in a mass. Magnetic resonance cholangiopancreatography (MRCP) may sometimes be used to evaluate segmental biliary ductal dilatation. Hepatic arterial phase imaging (on both MRI and CT) is especially important to determine whether a focal liver lesion is hypervascular. The differential diagnoses for a hypervascular liver mass are hemangioma, FNH, hepatocellular adenoma, HCC, fibrolamellar HCC, and a hypervascular metastatic lesion.
Recently there has been much interest in the use of the hepatobiliary contrast agent gadoxetic acid (Eovist/Primovist; Bayer Healthcare, Berlin, Germany) in evaluating focal liver lesions. Unlike conventional extracellular contrast agents, gadoxetic acid is transported into hepatocytes and subsequently excreted into the biliary system. This process is first evident at 2 to 3 minutes after contrast injection, with peak liver enhancement occurring between 15 and 20 minutes, the so-called hepatobiliary phase. The early dynamic images appear similar to those obtained with an extracellular agent. However, the normal liver becomes hyperintense at the hepatobiliary phase and any focal lesion that does not contain hepatocytes (metastases, cysts, hemangiomas, and so forth) will appear relatively hypointense. Lesions that contain normal hepatocytes, namely FNH, will appear hyperintense or isointense to surrounding liver. A small percentage of HCCs and hepatic adenomas can also demonstrate uptake of the agent. Gadoxetic acid–enhanced MRI has been shown to have higher sensitivity than CT and PET/CT in the detection of colorectal liver metastases. It is also highly useful in distinguishing hepatocellular adenomas from FNH (both hypervascular lesions). There is also interest in the use of hepatocellular agents for surveillance of small HCCs in the cirrhotic liver. The literature suggests that the use of gadoxetic acid allows for more accurate identification of small, early HCCs and high-grade dysplastic nodules in comparison with conventional extracellular agents. Disadvantages of MRI with gadoxetic acid include an increased incidence of dyspnea (and therefore respiratory motion artifact) immediately following injection, slightly less robust arterial phase, and decreased ability to evaluate for delayed enhancement (owing to hyperenhancement of the surrounding liver), a critical property of lesions such as hemangiomas and cholangiocarcinomas. Superparamagnetic iron oxide particle–containing contrast agents, which are taken up selectively by the reticuloendothelial system, have been used sparingly in the past and are no longer in routine clinical use.
Diffusion-weighted imaging (DWI) has emerged as an important pulse sequence in liver MRI. DWI characterizes the amount of diffusion of water molecules within tissues. Lesions with high cellularity cause restricted diffusion of water molecules and therefore remain hyperintense on DWI pulse sequences. These sequences are T2-weighted and, hence, lesions with high signal intensity on T2-weighted images may also appear hyperintense on DWI. In the liver, this technique is most useful to identify the presence and number of focal liver lesions (ie, lesion detection). However, there is overlap between the amount of restricted diffusion seen in benign and malignant liver tumors; therefore, caution must be exercised when attempting to characterize focal liver lesions with DWI in isolation. Along with fat-saturated T2-weighted imaging, it is one of the best sequences with which to identify lymph nodes.
Multidetector computed tomography
Given the high contrast and spatial resolution of multidetector CT (MDCT) and the ability to acquire isotropic data allowing routine multiplanar reformations in coronal and sagittal planes, in addition to its high sensitivity, widespread access, and speed of imaging, CT is widely used for the detection of liver metastases following the diagnosis of a primary extrahepatic tumor and for the imaging surveillance of these metastases following treatment. For the detection of metastasis, CT is performed only in the portal venous phase, typically acquired at 80 seconds following the start of the contrast injection. As with MRI, HCC surveillance using CT and characterization of an indeterminate liver lesion relies heavily on the dynamic appearance of the mass in arterial, portal venous, and equilibrium (delayed) phases of contrast enhancement. However, because CT uses ionizing radiation, there is concern for the risk of carcinogenesis, especially in younger patients. Multiphase liver imaging with CT thus has the downside of multiple exposures to ionizing radiation in a single examination. This fact, coupled with the superior sensitivity and specificity of MRI for characterization of liver lesions, relegates CT to the second-line modality. It remains an excellent alternative for patients who are claustrophobic, unable to maintain a 15-second breath-hold, or with other contraindications that preclude MRI.
For multiphase CT, the authors perform a reduced-dose unenhanced scan followed by intravenous contrast-enhanced dynamic imaging in the late arterial, portal venous, and mid-equilibrium phases. The unenhanced phase can be helpful for the detection of acute intratumoral hemorrhage, which typically appears hyperattenuating to the liver. The arterial phase is acquired 8 seconds after a threshold of 175 Hounsfield units (HU) is reached in the upper abdominal aorta, using automated bolus-tracking software. The portal venous phase starts 50 seconds after the start of the arterial phase, and the equilibrium phase starts 150 seconds after the start of the arterial phase.
Ultrasonography
Ultrasonography is not a first-line modality for characterization of a liver mass, and is not recommended by the ACR Appropriateness Criteria for this role. However, it can be a useful alternative for patients with renal insufficiency who are unable to receive contrast material on CT or MRI. It can also be helpful in confirming the cystic nature of a focal liver lesion or to confirm the diagnosis of a hemangioma seen incidentally on a portal-phase CT image or on unenhanced MRI. With a few exceptions (eg, a hemangioma and simple cyst), imaging features of liver masses are generally nonspecific on ultrasonography. When a lesion is detected on a sonogram, further characterization is best performed using contrast-enhanced MRI.
The American Association for the Study of Liver Diseases (AASLD) recommends ultrasonography as the most appropriate screening modality for detection of HCC in patients at risk. Ultrasonography has been reported to have a sensitivity of 65% to 80% and specificity greater than 90% when used for HCC screening. However, the performance characteristics of ultrasonography have not been as well defined in nodular cirrhotic livers undergoing surveillance. Patients with advanced cirrhosis undergoing screening for HCC often demonstrate marked heterogeneity of the liver parenchyma, which leads to limited ability to detect lesions.
General advantages of ultrasonography include no exposure to ionizing radiation, relatively low cost, and widespread availability. Limitations include high operator dependency, limited beam penetration in morbidly obese patients, and limited acoustic windows in some patients because of overlying bowel gases and rib shadowing. Research performed outside the United States on second-generation ultrasonography contrast agents has demonstrated high accuracy in characterizing liver lesions, but to date these agents have not been approved for hepatic imaging in the United States.
Nuclear medicine
Nuclear medicine imaging has a limited role in the modern-day characterization of a primary liver mass. Scintigraphic studies uniformly suffer from a low spatial resolution. FDG-PET/CT is the most widely used nuclear medicine examination, predominantly used in the setting of metastatic disease with an FDG-avid primary tumor. However, the high background uptake of FDG in the liver can limit detection of some lesions. Although most liver lesions showing increased FDG uptake are malignant, occasionally nonneoplastic infectious and inflammatory processes and benign liver masses can show increased FDG uptake. PET, however, has limited sensitivity for the detection of HCC (55%–64%), and even lower (18%) for cholangiocarcinoma.
99m Tc-labeled sulfur colloid may be used to distinguish between FNH and hepatocellular adenoma in patients for whom the glomerular filtration rate (GFR) precludes the use of CT or MRI contrast. Similarly, 99m Tc RBC scans can be used for the diagnosis of a hemangioma if the GFR precludes the use of CT or MRI contrast. Finally, an 111 In octreoscan can be a useful adjunct for the diagnosis or follow-up of patients with liver metastases from a neuroendocrine primary.
Malignant liver masses
Hepatocellular Carcinoma
Imaging features
Grossly HCC can present as: (1) a single mass, which can be large or small (satellite nodules may be present around the mass); (2) multiple masses scattered throughout the liver; (3) confluent small lesions; or (4) diffuse infiltration in the liver. Neoangiogenesis, characterized by formation of abnormal arterial supply, is an important trait of HCC. Large lesions tend to show central necrosis, and may contain intralesional microscopic or macroscopic fat. A capsule can be present in 65% to 82% of large HCCs, although a capsule can also be seen in large regenerative and dysplastic nodules. In general, HCCs associated with cirrhosis have a fibrous capsule, whereas those without cirrhosis are nonencapsulated. Venous invasion, particularly into the portal vein, is a characteristic of HCC, and the incidence of malignant portal vein thrombosis in association with HCC has been reported to range from 5% to 44%. However, patients with cirrhosis can also develop benign portal vein thrombosis secondary to portal hypertension and venous stasis, with a reported prevalence of 0.65% to 15.8%. A malignant thrombus is always contiguous with or directly in contact with a parenchymal tumor. Invasion into the hepatic veins or inferior vena cava can also occur, but is less common.
At present, ultrasonography and serum α-fetoprotein (AFP) are used to screen for HCC in cirrhotic patients. However, in some studies the sensitivity of these tests has been shown to be only 50% to 60%. If a lesion is seen on ultrasonography or the AFP level is higher than 20 ng/mL, further interrogation is warranted with a contrast material–enhanced MRI or CT scan. Large tumors in a cirrhotic liver, especially with vascular invasion, generally do not pose a diagnostic dilemma and can be readily diagnosed on CT. Unfortunately these tumors are associated with a poorer prognosis, and diagnosis must be focused on detection of small, potentially curable tumors. The detection of small tumors, defined as HCC measuring less than 2 cm, is challenging, however, especially on CT. The diagnosis of small HCC in a cirrhotic liver is confounded by a spectrum of nonmalignant nodules that occur in cirrhotic liver, ranging from benign regenerative nodules to dysplastic nodules.
On ultrasonography, tumors have variable echogenicity, and the most common appearance is that of a hypoechoic nodule (48%), followed by a mixed echogenicity nodule (25%). Contrast-enhanced MRI has greater sensitivity than CT, which in turn has greater sensitivity than ultrasonography for the detection of HCC in a cirrhotic liver. Pooled estimates of 14 ultrasonography, 10 CT, and 9 MRI studies for the detection of HCC showed a respective sensitivity and specificity of 60% and 97% for ultrasonography, 68% and 93% for CT, and 81% and 85% for MRI.
HCC has variable signal intensity on T1- and T2-weighted images, and the MRI signal may depend on the tumor size, grade, presence of intralesional fat, necrosis, or hemorrhage. Larger tumors tend to show greater heterogeneity. In general, tumors are hypointense on T1-weighted images and show a moderately increased single intensity on T2-weighted images. Sometimes tumors may be difficult to detect on T2-weighted images because of heterogeneity of the cirrhotic liver, which can obscure mildly hyperintense and isointense tumors. High signal intensity can be seen in some tumors on T1-weighted images and may be attributed to intralesional hemorrhage, fat, copper, or glycogen. Steatotic regions show loss of signal intensity on opposed-phase gradient-recalled echo images. Following contrast administration, on both CT and MRI the characteristic imaging feature of HCC is hyperenhancement (hypervascularity) during the arterial phase, with washout (tumor density or signal intensity lower than surrounding liver parenchyma) in the portal venous and equilibrium phases ( Figs. 1 and 2 ). In a patient with underlying cirrhosis, this combination of findings following contrast administration is considered diagnostic and sufficient to make the noninvasive diagnosis of an HCC. Washout of an arterially enhancing mass has a very high specificity for the diagnosis of HCC, with a reported sensitivity of 89% and specificity of 96%. Small HCCs can be isointense on T1- and T2-weighted images, and detected only in the arterial phase. On the delayed (equilibrium)-phase images, a thin enhancing rim (capsule) may be seen (see Figs. 1 B and 2 B). The arterial phase enhancement tends to be heterogeneous in large lesions and homogeneous in small lesions. The enhancing areas show washout in delayed phases. Large tumors may also occasionally predominantly show a pattern of intratumoral neovascularity in the arterial phase. Diffuse infiltrating tumors in a cirrhotic liver may be difficult to detect on unenhanced T1- or T2-weighted images, and can sometimes present a diagnostic challenge.
On postgadolinium images, diffuse tumors show patchy enhancement in the arterial phase, but washout is seen in all tumors on the delayed postcontrast images. Furthermore, they are frequently associated with portal venous tumor thrombosis and elevated serum AFP levels. Subtraction of the unenhanced from the enhanced MR images is extremely helpful in assessing enhancement in nodules that have high signal intensity on unenhanced T1-weighted images. On gadoxetic acid–enhanced MRI the tumors typically show a low signal intensity in comparison with the surrounding nonneoplastic hepatic parenchyma in the hepatobiliary phase. Malignant tumor thrombus typically shows increased signal intensity on T2-weighted images and expansion of the vein, compared with a low T2 signal intensity and near normal caliber of the vein in bland thrombosis. Following contrast administration on CT and MRI, the presence of neovascularity or heterogeneous enhancement within the vein on the arterial-phase images is considered highly specific for tumor thrombosis ( Fig. 3 ).
Based on the current recommendation of the AASLD, biopsy is not needed if a mass greater than 2 cm in diameter shows classic features of HCC (hypervascularity in the arterial phase and washout in the venous phase) on either CT or MRI or when a mass 1 to 2 cm in diameter shows these features on both CT and MRI. Therefore, imaging plays a crucial role in the diagnosis of HCC in cirrhotic livers. However, arterially enhancing nodules less than 2 cm in diameter are not uncommon in the cirrhotic liver. These nodules can pose a diagnostic challenge and dilemma, because most are benign and represent either regenerating and dysplastic nodules or arterioportal shunts, but those that represent small HCCs have a chance for curative treatment. Nodules that show only arterial-phase enhancement and no washout in the portal venous or equilibrium phases are considered indeterminate and cannot be characterized as definite HCCs. Such nodules need closer follow-up imaging or, in challenging cases, biopsy. The United Network for Organ Sharing has recently added rapid growth as a diagnostic criterion for HCC for nodules with a diameter greater than 1 cm at initial diagnosis and showing hyperenhancement in the late arterial phase, although this criterion should be used with great caution. In cases of doubt, short-term follow-up imaging or biopsy may be considered. It is also important to distinguish such nodules from nontumorous peripheral or subcapsular wedge or geographic areas of arterial-phase hyperenhancement, referred to as transient hepatic density (on CT) or intensity (on MRI) differences. Radiologic criteria favoring HCC are size larger than 2 cm, moderate hyperintensity on T2-weighted imaging (as opposed to dysplastic nodules, which are typically hypointense on T2), washout in the delayed phase, enhancing tumor capsule on delayed images, and rapid interval growth.
Liver imaging reporting and data system
The Liver imaging reporting and data system (LI-RADS) was developed to standardize the terminology, interpretation, and reporting of HCC and other focal liver lesions in patients with cirrhosis or other risk factors for HCC. The latest version (v2013.1) applies to CT and MRI performed with extracellular contrast agents. Hepatobiliary contrast agents are expected to be incorporated into upcoming versions. Focal observations in the liver are assigned an LI-RADS category from 1 to 5, with increasing likelihood of being HCC (see Fig. 2 ). Importantly LI-RADS category 5 lesions are considered 100% certain to represent HCC, and this category is essentially equivalent to the Organ Procurement and Transplant Network class 5. Observations are assigned to categories 3 through 5 based on the presence or absence of arterial hyperenhancement, size, and major criteria (washout appearance, capsule, and threshold growth), as shown in Fig. 3 .
Fibrolamellar HCC
Fibrolamellar HCC is a rare primary liver tumor that predominantly occurs in young adults, with most cases occurring in patients without underlying hepatitis or cirrhosis.
Imaging features
Fibrolamellar HCC typically presents as a large, well-circumscribed, lobulated solitary mass in a liver without cirrhosis, histologically characterized by the presence of fibrous strands in the tumor. A central stellate scar has been reported in up to 71% of cases. A capsule may be present, although it can be incomplete. Focal calcifications are common and can be seen in 40% to 68% of cases ( Fig. 4 ). As opposed to conventional HCC, portal vein thrombosis is uncommon in fibrolamellar HCC and occurs in only 5% to 10% of cases.
